Microhabitat use by freshwater mussels and recommendations for determining their instream flow needs

A conventional application of the instream flow incremental methodology (IFIM) assumes that target organisms have specific microhabitat preferences and the ability to move to areas of suitable hydraulic conditions in response to changes in stream discharge. We investigated the use of the IFIM for determining the instream flow needs of a diverse mussel assemblage in Horse Lick Creek, a fourth-order stream in the upper Cumberland River drainage in Kentucky. We determined habitat availability by measuring water depth, velocity and substrate at 60 cm intervals along 23 transects during low, medium and high flows. The distribution of mussels within the study site was highly contagious. Although habitat suitability curves developed from data collected on 2004 mussels indicated a clear preference for particular hydraulic conditions, the limited mobility of mussels in the coarse substrate of Horse Lick Creek implies that these curves are flow-conditional—that is, mussels appear to prefer different hydraulic conditions at different stream discharges. Consequently, these curves are of limited value for determining conservation flows for mussels. Nonetheless, water depth and velocity were important factors limiting the distribution of mussels during base flow periods. Similarly, substrate characteristics were of limited value in defining mussel distributions; unfractured bedrock excluded mussels from portions of the study site, but mussels did not utilize all areas with preferred substrate. Because the larvae (glochidia) of mussels in Horse Lick Creek are obligate parasites on fish, data were also collected on habitat preferences of the host fishes. These data were incorporated in the physical habitat simulation system (PHABSIM) to determine the relationship between the availability of host fish habitat and stream discharge during periods of glochidia release and juvenile settlement. Unlike simple hydraulic variables, complex hydraulic characteristics such as shear stress were significantly correlated with mussel abundance for flows ranging from 0.03 to 2.18 m3 s−1. This range encompasses most flows during the period of juvenile settlement. We suggest that the high shear stress in some portions of the study site is a major factor limiting mussel recruitment. The lack of a significant correlation between mussel abundance and shear stress at high flow (9.35 m3 s−1) resulted from a variable relationship between shear stress and discharge among transects due to channel morphology. The higher shear stresses at most transects over mussel beds during a discharge of 9.35 m3 s−1 suggests that spates occurring during or shortly after juvenile settlement may result in a loss of juveniles. The unique life history and limited mobility of mussels necessitates a more complicated procedure than generally used for fish and other macroinvertebrates for determining conservation flows. Specifically, we recommend an approach that incorporates concepts of hydraulic stream ecology with the more common practice of modelling only simple hydraulic variables in habitat simulations. Estimating the complex hydraulic key characteristics can be performed with minimal effort through the selection of appropriate subroutines with PHABSIM. This approach may also be suitable for simulating habitat of other sessile organisms.