BIOSCAN and Design of Field Collections

BIOSCAN and Design of Field Collections

The BIOSCAN program aims to promote significant field-based collection of specimens and samples of different organism groups from all ecosystems. This thread is for general discussion of optimal design of collection strategies to maximise the value of the information derived from these efforts. If appropriate, subtopics will be broken out into separate threads for more focused discussion. This opening post will be maintained as an introduction and overview of discussion topics.

Discuss experimental design for BIOSCAN field collection activities below.

Need to address a variety of contexts

Many opportunities exist globally to expand coverage for barcode sequences and sample new ecosystems. In many of these cases, it will not be possible or appropriate to require standard or complex methodologies to be followed. The Kruger Malaise Program is an excellent approach to sample widely across an important area, but this program relies on busy rangers looking after malaise traps as part of their weekly schedule. Citizen science and community groups are likely to be valuable BIOSCAN partners in many countries but may need simplified protocols that can be learned and followed easily. Existing ecological monitoring networks may be able to include collecting activity for BIOSCAN at their sites, but these are often in remote locations and visited less frequently than the weekly schedule sought for BIOSCAN malaise trapping.

On the other hand, there will be significant information gain if we are able to design good protocols for selecting sampling sites and for gathering informative metadata elements and associated measurements at each site. Recording a standard suite of site variables (soil type, aspect, water pH, etc.) or collecting a standard set of complementary samples or observations (malaise trap, pitfall trap, vegetation survey, plant phenology data, etc.) may facilitate valuable ecological research and wider inclusion of BIOSCAN as an element in monitoring schemes.

For such reasons, BIOSCAN needs a strategy that includes at least two different thresholds for sampling activities, a well-defined minimum protocol that can be cheaply and easily applied in many situations and an optimal protocol that includes steps both for planning sampling efforts and for collecting associated data and metadata. It may be appropriate in some situations to define more than two distinct thresholds, provided that there is a clear understanding of the benefits derivable from each additional level of effort.

I would be interested in exploring the possibility of running a number of malaise traps in different ecosystems with a detailed plan for capturing significant associated metadata and samples/observations of other taxa. For example:

  • Metadata on trap placement
    • Latitude, longitude, altitude
    • Orientation
    • Slope (in multiple axes)
    • Vegetation height
    • Distance to edges
    • Distance to water
  • Metadata on climate/microclimate
    • Weekly rainfall
    • Weekly temperature range
    • Wind directions and speed
  • Observations of vegetation
    • Vegetation survey within fixed radius (or radii) around trap
    • Flowering and fruiting phenology
  • Sampling and (meta-)barcoding of other groups
    • Soil biota
    • Pitfall, pan or light trapping

A good sample of metadata and coordinated observations of this kind (including for some traps positioned along gradients or in close proximity) would be likely to give useful information to determine which metadata elements are most informative for interpreting variation between trap samples. These elements could be incorporated into best-practice recommendations for standardised BIOSCAN malaise trapping.

Clearly, the questions implied by this example are ones of much wider general interest to ecologists and it is likely that we could learn much of relevance from past experiments and published studies.

Another important topic is the metrics/measurements/data by which we will assess any of these experiments.
One traditional metric is COI quality/length after PCR, but one might now also use a direct measure of whole genomic DNA quality.
Good barcodes are the goal, but perhaps there are metrics upstream of PCR for particular markers that would predict success for an array of markers, rather than assessing each one?

Thanks, Jonathan - can you point to any studies that seek to follow this approach?