For our purposes, the species-specific percentage cover of plants, coupled with some measures of seed productivity, will be sufficient to discern temporal and spatial differences in plant food resource abundances. To accomplish this:

1. Establish a total of 16 one-square meter quadrats in association with each trapping web (or grid). With generally 2-3 replicate webs per site, this would create 32-48 samples for vegetation. The plant quadrats can be establishedin clusters of four quadrats/cluster, arranged in a square, with 10 meter intervals between quadrats. Each of the four clusters (with 4 quads each) should be located at least 20 meters outside the web perimeter to avoid trampling by rodent trapping crews. We would suggest orienting the 4 clusters of quadrats in a standard fashion, such that clusters occur 20 meters outside of traps numbers 12, 48, 84, and 120 (cardinal directions from the center of each web).

2. Each quadrat should be permanently marked with rebar in at least two (opposite) corners, so that the sampling frame (described below) can be relocated exactly each time the quad is sampled. Quadrats should be numbered with permanently attached metal tags so that data from various sample periods refers to the same quadrat.

3. Quadrats should be measured with a 1 x 1 meter PVC-pipe frame, with strings criss-crossing the frame at 10 cm increments. This will create a 100-grid square frame, making it very easy to determine percentage cover of plants.

4. To sample the plants on the quadrat, place the frame on the plot, but taking care not to crush the vegetation. For some taller vegetation types, it may be necessary to build the frame with PVC "legs" (like a small table) that would keep it above the ground and not press or damage the plants. When the frame is in place, the observer looks straight down on it, and estimates the amount of cover for each species of plant present; this is done by counting the number of 10x10 cm squares with a particular species of plant in them. Partial squares can be tallied as well, if the plant does not cover the entire square. Each full 10x10 cm square equals 1% cover of the entire quadrat.

NOTE: Most quadrats of this size usually have fewer than 10 plant species, and thus the data collection of the plants takes on the order of 4-6 minutes per quadrat; there will likely be some slower times at first, until the field crews learn the identifications of the common plants. The entire plant data collection on up to 48 quadrats could be done by a team of two people in 2 days utilizing the "middle" part of each day (assuming rodent trapping in the mornings and rebaiting in the late afternoons).

5. We suggest sampling plants two times per year; once in the early part of the summer (to estimate the cover of spring plants), and once late in the summer to get the full summer's production.

6. Seed production: At the time that plants are being sampled, estimates of seed abundances can be made. These can be more qualitative than quantitative; i.e., seed numbers per plant of each species can be measured as:  NONE PRESENT

1-10 SEEDS OBSERVED PER PLANT

10-50 SEEDS OBSERVED PER PLANT

50-100 SEEDS OBSERVED PER PLANT

> 100 SEEDS OBSERVED PER PLANT

Such data will likely be able to distinguish between drought years (practically no seed production) and wet years (loads of seeds).

7. The data from all of the quadrats can be averaged by species or pooled to get total plant cover and seed production. We can develop a common data sheet if everyone is interested in doing this (probably a good idea...)

8. Some estimate of juniper berry, pine nut and acorn production should be undertaken in the fall of each year for those sites that have such species. We are currently working out a method to do this, and should be able to provide you with protocols later this summer.

9. Plant coverage should be subdivided into grasses, herbs, and shrubs, then analyzed calculated for each site. Changes in cover can then be calculated through time and correlated with rodent abundance changes.