The GLUSEEN pilot experiment selected three (3) of the six (6) habitat types presented in Table 1 in this paper. The habitat types are a combination of level of management (M) and disturbance (D), each being low (L) medium (M) and high (H). Disturbance should be evaluated both above and below ground, as described below. These categories are descriptive and thus qualitative; moreover, they will be relative to the climatic conditions of the city location. We followed this protocol for the pilot project and hope that the criteria described below will help to achieve a uniform site selection procedure and thus comparable results.
Our habitat types located inside the city were:
- LD/LM: ‘Remnant’
- HD/MM: ‘Turf’
- HD/LM: ‘Ruderal’
- ‘Reference’, which is located outside the city
Protocol for Habitat Selection
- Start with a soil / geologic map, if available. It is important to keep the soil type or geologic substrate (parent material) as similar as possible.
- If the underlying material is spatially uniform, no further investigation is needed
- If the parent material/soil is a patchwork and you are forced to select more than one soil type try to use similar soil types. For example, in the U.S., a similar soil type is either a competing series or geographically associated soil. Competing series are members of the same family, which includes particle size, mineralogy, and cation exchange capacity. In this case make sure that at least one soil type will have all four habitat types.
Note: The soil type choice depends on the number of suitable ‘Remnant’ habitats (see below). We expect this to be the limiting factor within urban landscapes.
- Habitat types
- ‘Remnant’ (LD/LM): Select these patches using a vegetation map, if necessary. The ‘Remnant’ should reflect the vegetation and soil type of the ‘Reference’ as closely as possible, although it may differ in slope, soil type, or site history (e.g. secondary forests on former agricultural areas). It may have a somewhat different species composition and understory than the ‘Reference’. Do the best you can. To confirm that the soil was not physically disturbed below the surface layers, dig down 30-50 cm with a shovel or soil probe and determine if there is natural soil horizonation, i.e., is there an A horizon followed by a horizon that is typical to the area, such as a B horizon. In many cities plant cover is highly altered, thus the number of ‘Remnant’ patches might be the limiting factor in habitat selection. In this case you may be forced to minimize other possible confounding factors by selecting the remaining habitat types close to the ‘Remnant’ patches.
- ‘Turf’ (HD/MM): This will be a grassy area and it includes public green spaces and lawns, but not high maintenance lawns such as golf courses. We define maintenance as supplementing with fertilizer, lime, or herbicides. The level of irrigation will depend on the regional climate, specifically precipitation. We assume all turfgrass systems are mowed/clipped. Suggested preference: parks > residential lawns (but see patch size below), as long as they are not of high maintenance > turfgrass on campus > commercial turf.
- Ruderal’ (HD/LM): These patches will have highly disturbed soils with no sign of natural soil development throughout the top 30-50 cm of the profile. Typical to these sites are the existence of human artifacts such as building rubble or refuse, etc. Examples include recent or ongoing construction sites and vacant lots. These habitats usually have little or no overgrown vegetation, and soils with low organic matter content. Since soils and the aboveground vegetation can be disturbed many different ways, ‘Ruderal’ habitats may differ even within the same city. Time since disturbance may also play a role.
- ‘Reference’ should reflect the regional biome as determined by climatic patterns and soil development. An example would be a deciduous forest in the Northeast U.S. The selection should be determined by distance from the city, so that no urban environmental effects are present.
- Additional guidelines for patch selection
- Avoid riparian areas and other poorly drained areas. Often, these are green corridors with dense tree cover that may appear as ‘Remnant’. However, conditions there will be very different from upland forest stands.
- ‘Remnants’ can often be found on the top of hills where construction limitation prevents turning lands to residential areas. Again, these should be avoided if possible.
- A patch should be at least 30 x 30 m in size, preferably larger.
- Ideally there should be five replicates per habitat type. Once these have been selected, the experimental plots have to be established. Plots for the teabag deployment and additional soil sampling are 2 m x 1.5 m in size. In this area establish a 5 x 4 grid with each grid point being 50 cm apart. Bury one teabag at each grid point as described in the teabag protocol (Supplement 2 in this paper). Soil sampling and soil biota sampling should take place around the perimeter of this grid. Note: sampling should be a minimum of 10 m from the edge of the habitat patch, preferably more if the edge is a road. Around 30-40 m would be ideal, but often the patches themselves are smaller than that.
- Terminology to be used in your logbook and datasheet:
- ‘Site’ is the city under study
- ‘Habitat’ or ‘habitat patch’ represent the habitat types we described above. Distinguish the replicates by a number e.g. Turf1, Turf2, etc. Additionally you may want to use a specific name (Linkwood Park, Campus lawn)
- ‘Plot’ is the area within a given habitat patch where the teabags will be deployed, and soil sampling and monitoring will occur.