General Tips

Figuring out how many samples to take is not easy, too few and you will not find differences, too many and you spend an eternity with sample processing. Experience with a system, and establishing large enough treatments can be passed down from colleagues and advisors, and this may be one of the best ways to choose sample sizes. Power analysis can also be used if you can guess variability. Last a pilot study can help, figuring out if your intuition and calculations are serving you well before a large field campaign begins.

Plant Sampling

Leaf Sampling

Leaves are commonly sampled as they contain the photosynthetic tissue, and much of the nitrogen in a plant, and thus can be useful metrics of nutrient dynamics, photosynthetic capacity, and can also be measured for specific leaf area, which indicates plant investment in photosyntheic area compared to biomass, a genetic metric of ecophysiological strategy. For evergreens leaf sampling can be more tricky than deciduous trees as needles can stay for several years. New growth is typcially easy to identify, at the ends of branches and is usually more bright green. Sampling new growth gives a certainty of the time, but can be nutrient rich compared to older needles, so comparisons need to take this into account. Conifer needles are also more challenging to measure specific leaf area because of greater 3D structure.

Above/Below Ground Biomass

For smaller plants above and below ground biomass are useful general metrics of plant ‘investment’ strategies, sampling the entire plant shows whether more carbon has been directed towards nutrient and water acquisition or towards photosynthesis. This can be useful for interpreting other data, such as soil quality, or if in a managed system, may help understand the efficacy of treatments across different sites.

Tree Rings

Generating a chronology of a plant’s growth patterns is possible in temperate trees which produce seasonal growth rings. I cannot cover this topic extensively here, however will comment on its use. Relatively inexpensive tools can be used to quantifiy hundreds of years of tree growth, and with strategic experimental design, many interesting things can be learned about growth pattern and response of trees. Tree rings are collected using incrememnt borers which remove a small cylinder, or tree core, that can then be used to count growth rates over time. Skilled researches can do so with exact annual accuracy, allowing for specific correlations and reconstructions with other historically available data, and can generate a site history that informs on precipitation, fire, flood, avalanche and other natural occureneces.

Soil Sampling

How deep?

For convenience soil is commonly defined as the first 2 meters of material below ground (or less if there is shallow bedrock). If you are intersted in classifying a soil according to the keys of soil taxonomy, a more elaborate excavation must be conducted. I’ll not be giving advice on that front, rather I this is aimed to help more targeted sampling efforts.

How many?

Soil is very heterogenous so expecting low effect sizes over space, and between treatments is a good place to start. See my discussion on a priori tests of statistical power to help decide how many samples you need. Suffice to say, you probably want more than the 3 required to make a standard deviation!

Active Biological Zone

Commonly people will sample the top 10-15 centimeters of soils for microbial, nutrient, and carbon analysis. The reason is that this depth has been determined as the most active zone for biological cycling, thus different ecosystem types, or imposed treatments that may impact soil carbon, nitrogen, phosphorous cyling, may be most likely to generate an impact on this zone. Naturally, this will have an impact on the microbial community responsible for that cycling, and potentially on associated soil physical properties (e.g. water holding capacity, infiltration).

Rooting Zone

If you want to go a little deeper and understand the impact of soil on water availability at a site, you may want to do some research as to the rooting depths and strategies of the plants in the area. If big trees are involved, naturally you can’t excavate or auger many meters down, but as clay content of a soil is largely responsible for its surface area, and resultant water holding capacity, it is worth augering down as close to 2 meters as possible since most of the clay is likely in that zone, and it may be a good proxy for understanding the overall soil water capacity. Identifying depth increments is then required, and should be done as an educated guess (e.g. in the U.S.A. check out soilweb, which is also available as an app), or by digging a real soil pit, identifying depths of important soil horizons, and then sampling accordingly.

Impacts of topography

It is worth thinking about how topography impacts soil sampling and either controlling for it, or sampling many locations within the local topography. The top of a hill tends to be an eroding area, while the valley, or gulley below tends to be a place of accumulation. This also relates to infiltration of water based on that surface topography, making for differing rates of pedogenesis within short distances. For example, the top of a hill to the bottom flat area, could have shallow rocky, and deep clayey soils as a result of this weathering gradient. For a brief discussion on the topic, check out the wikipedia page on ‘catenas’, the formal name for this landform and concept.

Exploratory Sampling

Commonly I have sampled 0-15, 15-30, 30-60, 60-100, 100-150, 150-200 cm in soil profiles. This has served me well in cases where I was just trying to get an idea of a site, and generate some exploratory data. It catches the surface horizons, clay rich horizons, and helps estimate soil depth (do I hit bedrock, or consolidated bedrock at or before 200 cm?).