Hello, I am confused with the design of experiment and I need your help. I study macroinvertebrates from 40 lakes. Each lake was sampled employing stratified random technique. Each lakes was divided into 3 different lake zones based on light penetration: near-shore area, transitional, and the deepest part of the lake. 5 random sediment samples were taken from the near shore area, 5 random samples in the transitional zone, 25 random samples in the deepest part of the lake. Sediment type and organisms' counts was recorded for each sample. To complicate more, 2 different types of equipment were used. Near-shore and transitional area were sampled with large-area heavy sampler, but the deepest part of the lake was sampled with small area sampler. I do not know if different type of equipment should figured in the model since all organisms counts were calculated as per sq/m. I want to see if there is the sediment effect on organism counts if lake and zone are taken into account.I consider the lake to be the random effect . Zone is the meaningful, it carries information for benthos communities. Zone is nested within the lake. Sample is nested within the zone and lake. One sample - one sediment type. What is the correct way make the model: My guess is something like: aov(Count ~ Sediment+Error(Lake/Zone/Sample)) Sincerely, Anna [[alternative HTML version deleted]]