This study investigated the main causes of population abundance fluctuations. Particularly, attention was paid to whether a density-dependant factor, such as a stock-recruitment relationship (S-R relationship) or a density-independent factor such as an environmental factor, is more important. Using data pertaining to the number of eggs of the Pacific stock mackerel and information about regime shifts and sea surface temperature, the shape of the S-R relationship was discussed and these shapes with the results of simulation trails were compared. Further other historical S-R relationship data were analyzed. The results are as follows: (i) a new mechanism that causes population fluctuations could be proposed, that is, (a) the recruitment is proportional to the spawning stock biomass (SSB) and the relation is expressed by several lines with the same slope, and (b) the shift between the lines occurs due to environmental conditions; and (ii) the density-dependent S-R relationship, which suggests that recruitment decreased due to high density in SSB, proposed by Ricker or Beverton and Holt, would not exist.KEY WORDS: stock-recruitment relationship, density dependence, mackerel, regime shift, sea surface temperature, Ricker model, Beverton and Holt model.