Watershed biological information flow (WBIF) is defined as the path, processes and control of biological information transport, exchange, interaction and feedback among different spaces and systems along with watershed ecosystem processes, and could be partly described as the land-to-river and upstream-to-downstream bioinformation transportation (including organisms, nucleic acids, peptides and other biomarkers), which is driven by the hydrologic processes of watershed systems. The WBIF labels the transport of organic matter and energy. The WBIF integrates the ecological processes of environmental DNA (eDNA), including the origin, state, transport, and fate of eDNA, and makes it possible that the species composition in river system is monitored and assessed using eDNA. The WBIF estimation is the key for watershed ecosystem processes studying and riverine biodiversity monitoring. However, in practice, the parallel samples in each sampling site always are limited. And how parallel samples would impact WBIF estimation is unknown. Based on the principles of stochastic sampling survey, we hypothesized that parallel samples would not impact the accuracy of the WBIF estimation, but affect the precision of the WBIF estimation. Then, we transformed this hypothesis into a set of formulas and tested it with a series of analog computation. Results showed that the number of parallel samples (efficiency of detection) affected both the accuracy and precision of the WBIF estimation. The optimal WBIF estimation was less than the actual WBIF in any condition. Along with the increase of parallel samples (efficiency of detection), the optimal WBIF estimation gradually neared to the actual WBIF, the range of WBIF estimation gradually focused on the actual WBIF. In other words, more parallel samples (higher efficiency of detection) led higher accuracy and precision of the WBIF estimation. In addition, the actual WBIF affected both the accuracy and precision of the WBIF estimation too. Larger actual WBIF led higher accuracy and precision of the WBIF estimation. The relative relationship between the number of biological information types in upstream and downstream samples affected both the accuracy and precision of the WBIF estimation too. The accuracy and precision of WBIF estimation would be higher when the number of biological information types in upstream samples was more than those in downstream samples. So, we suggest that in the work of watershed ecosystem processes studying and riverine biodiversity monitoring, the relationship between parallel sample number and detection efficiency should be assessed, the suitable parallel sample number should be estimated based on the reliability target of WBIF estimation, the sampling program should be designed with suitable parallel samples, the WBIF should be estimated based on all parallel samples of each sampling site, at last the estimated results of WBIF should be re-evaluated according to the posterior probability of WBIF in different conditions. The current work provided the framework and methodology reference for the post-evaluation.