Marking razor clams for Mark Recapture Method

To overcome this obstacle, a modified version of the Peterson Mark Recapture method was used for many years. This method required digging hundreds of clams and then marking and "planting" them back into specific areas (giving scientists an estimate of the number of clams that show at any given time). Then two weeks later biologists returned to redig all clams that would "show" in a specific area, both marked and unmarked. The large number of problem variables in this method lead biologists to begin looking for a more reliable technique. In early 1992, Dan Ayres, a Washington biologist heard a presentation in Canada on a new technique being used in Cook Inlet by the Alaska Department of Fish and Game. With the support of agency administration, two members of Washington's razor clam staff made a trip to see first hand how this technique worked and to determine if it could be modified for use on the Washington coast. Starting the summer of 1993 and going through the summer of 1996, this new method was modified and tested in a variety of situations and conditions. After many long hours of work, it was determined that this method was capable of producing more accurate population data. So, starting in 1997, it became the method of choice for determining the total number (abundance) of razor clams on each of the five beaches managed by Washington State Department of Fish and Wildlife (WDFW). This new method is referred to as the Pumped Area Method.

A typical Pumped Area sample transect.

Pumped Area Method

To determine the total number of razor clams on a given beach, it is necessary to have an accurate estimate of two things: the average density of razor clams (number of clams per square meter) and the total area of the beach (number of square meters). The Pumped Area Method gives a good estimate of razor clam density by estimating the number of clams in sample plots of a known area ( ½ square meter ), averaging the number over the length of one transect on the beach and then expanding that density to the total survey area.

The Random Sample

As we begin the task of determining the abundance (total number of razor clams) on any of our management beaches, it is critical to be able to assume that every square meter of the beach is equally likely to be sampled. To meet this assumption, we randomly select each transect location. Upon arrival at the transect chosen for a given day's work we begin by locating the upper limit of the clam bed. We determine the location of our first elevation (a line perpendicular to the transect) by choosing a random number between one a fifty. We then measure down (from the top of the clam beds) that number of feet to establish the first elevation. Every subsequent elevation is then located in fifty foot intervals. Each elevation runs either north or south of the transect and this too is selected randomly. At each elevation we pump a total of six plots.

Using a 5 hp water pump to deliver water from the surf, clams within a 0.5 sq meter area float to the surface.

The Data Gathering

Determining the number of razor clams in each sample plot is the key data point the pumped area is designed to collect. Clams are removed from the sample plot by liquefying the sand with a high volume jet of water which causes the clams to float to the surface so they can be gathered, counted and measured. This is achieved using a 5 hp water pump (producing 60 gallons/minute) to take water from the surf, or a nearby lagoon and moving it as much as 600 feet up the beach to the desired location. The water is pumped into the plot surrounded by aluminum ring. Each plot is pumped for exactly 3 minutes, although most of the clams float to the surface in the first minute. We assume that every clam in the area of the plot is removed and counted. The water pumped into each plot liquefies the area down to 48 inches. Our experience shows us that razor clams do not dig beyond that depth.

The clams float to the surface as the sand is liquified.

As the plots are pumped at each elevation along the transects, our crew carefully records the number of clams found in each plot. These clams are then measured using our standard cork lined measuring board with a sheet of water proof paper attached. The clam shell length is "punched" onto the sheet. Back at the lab a digitizer connected to a desk top computer is used to "read" the punches with the edge of the sheet as a reference point. Each "punch" represents the length of one clam. Clam lengths are recorded by date, transect, elevation and plot. After the clams have been measured and recorded they are returned to the plot unharmed, (occasionally a clam will be damaged by the force of the water).

Fish Biologist Dan Ayres records number of clams per plot per elevation and the length of each clam. To demonstrate how deep a plot is following pumping, Scientific Technician Clayton Parson (wearing waders) steps into a freshly completed plot.

One full transect takes five hours to complete. As the tide changes, the completed plots quickly fill with sand. By the next low tide, no sign of the previous day's activities are visible.

Data Analysis

The data collected is later used to determine the population of razor clams on each beach. The recorded clams are sorted by size, all clams 75 mm and under are considered pre-recruits. These smaller clams are less likely to be harvested by diggers and are not used in the calculation of the total allowable catch (TAC). Clams over 75 mm are considered recruits and available for harvest. The clam totals are entered into a computer model by transect, elevation, plot and by category (re-recruit or recruit). The model determines the average density (clams per square meter) for each category and then expands by the total number of square meters of razor clam habitat on each beach. As a result, estimates of the total number of all clams, the total number of pre-recruit clams and the total number of recruit clams, are calculated for that beach. The total number of recruit sized razor clams is then used in a formula that calculates the TAC.