Looking for a method to get better grasshopper surveys

[Bruce McKinlay]

A starter for ten.

Central Otago DOC are reviewing the Siguas surveying and monitoring programme they manage. 

These are small cryptic grasshoppers found on gravel stone heaps

Currently surveys are undertaken on 30m x 30m plots which are subdivided into 5m x 30m strips.  Each strip is walked by a group of people and grasshoppers are counted along the way.

There are unresolved issues with observer skills and bias and also the survey is quite suseptible to change in environmental conditions.  In other words if the sun goes behind a cloud then the survey becomes null and void.

My question is: Are you aware of a more consistent or effective grasshopper monitoring method.   

Thanks very much for any offerings

Bruce

[Murray]

I don't know Sigaus, so apologies if this doesn't hit the mark...
I would guess it's mostly a matter of doing better what you do already - improving the visual count index (assuming you don't want to get into expensive capture-recapture). How to do it better?

1. Collect detailed data on covariates that will influence detectability (sun, temperature, observer etc.) and use them in a formal model to standardize the index
2. Be nice to experienced observers and make sure they come back
3. Deliberately rotate observers around plots and resample plots with different observers within a year (according to a formal scheme)
4. Think about a better spatial sampling scheme - placing the 5 x 30 m strips randomly within the larger area you want to survey, where 'randomly' can get quite sophisticated (stratified, or more spatially representative 'generalised random tesselation stratified' etc.)
5. Think about a better temporal sampling scheme (I'm assuming you sample the same sites every time). It can be better to schedule replacement of a fraction of sites each year.

Hope this helps some