Because of the high management relevance, commercial fish related aspects have often been central in marine ecosystem investigations. The iterative shiftogram method was applied to detect occurrence, type and timing of shifts in the single and multivariate time series linked to the spring spawning herring larvae in the Gulf of Riga (Baltic Sea). Altogether nineteen larval herring and related environmental variables were utilized during the period of 1957-2010. All the time series investigated, either single or multivariate, exhibited one or more shifts with variable type and timing. Multivariate shiftogram based on all time series identified two distinct states (1957-1983 and 1992-2010) in studied variables, separated by a smooth transition period lasting almost ten years. The observed shift was mainly related to hydroclimate and not to phenology or biota. Significantly increased variability was found in larval herring and recruitment abundances after the shift. While the shift in hydroclimate (1985-1991) was followed by the shift in phenology (1991-1997), the shift in biota occurred remarkably later (2003). It is likely that the dynamics in biota were affected by other drivers than those investigated in the current paper.