Multiple Tube Sampler for Benthic and
Pelagic Invertebrates in Shallow Wetlands

We prepared samples for shipment and storage by using an improved self-cleaning screen (Euliss and Swanson 1989) to remove sediment particles from both water column and benthic fractions of the samples. With the sediment core held in place, the pressure plug was removed and the water column fraction of the sample was decanted into the self-cleaning screen. Sediment cores were forced partially out of the sample tubes by shaking, so that a little of the core extended from the tube. The protruding section was pulled by hand to avoid damaging fragile invertebrates, and aligned with a permanently etched scale on the clear tube that facilitated visually precise sampling of core profiles. Core material that extended out of each tube was trimmed flush with the end of the tube leaving a sediment core of known thickness within the tube. Cores from individual tubes were processed in an improved self-cleaning screen (Euliss and Swanson 1989) to remove soil and fine organic debris. Water column and benthic samples were stored in separate specimen jars because water column invertebrates are often translucent and difficult to sort when mixed with organic debris that is characteristic of benthic samples.

We evaluated our multiple tube sampler from November to January 1986-87. We collected invertebrate samples from a seasonal and a semi-permanent wetland on the Sacramento National Wildlife Refuge in Willows, California with the multiple tube sampler, a modified water column sampler, and a standard benthic sampler (Swanson 1978a,c). The water column sampler we used had been modified by adding a metal door to seal the bottom of the tube just before samples were extracted from a sampling site. The seasonal wetland was a moist soil impoundment managed for swamp timothy (Heleochloa schoenoides), and the semipermanent wetland was managed for submersed aquatic plants. Vegetation in both ponds was senescent and the ponds had little or no emergent cover. The seasonal wetland contained an 8- to 10-cm thick mat of decomposing swamp timothy at the benthic-pelagic interface. Vegetative debris in the semipermanent wetland was largely decomposed and the benthic-pelagic interface was devoid of vegetative debris. Both wetlands had flat bottoms and water depth was approximately 25 cm.

We sampled ponds biweekly, and each pond was sampled 6 times. Each time a pond was sampled, 5 random locations were selected, and a multiple tube, a water column (Swanson 1978a), and a benthic (Swanson 1978c) sample were collected at each location. Individual tubes of the multiple tube sampler were kept separate to facilitate sorting, but the results were pooled into a single sample prior to statistical analysis. Likewise, water column and benthic samples collected with separate sampling devices were sorted individually, but the results were pooled into single samples to facilitate comparisons with the multiple tube sampler. A self-cleaning screen (Euliss and Swanson 1989) was used to clean all samples, regardless of sampler type. Invertebrates were identified using Ward and Whipple (1959), enumerated, and density recorded as number of individuals per m². To normalize the data and stabilize variance, invertebrate densities were transformed to their natural logs. Because natural logs of zeros cannot be taken, we added 1 to each observation prior to transformation. A repeated measures ANOVA was used to evaluate the effect of wetland and sampler type on invertebrate densities.