For occupancy models, as you know, it is generally recommended that grid sizes should be bigger than an animal's home range. In this case, it would probably be okay to consider the mean home range, as snow leopards display different home range (large and smaller home ranges mainly depending on prey availability). With this you could go for the SECR models.

Occupancy models can be used to look at 'habitat use' as well 'true occupancy' with presence-absence data. For habitat use or intensity of space use, the above assumption to make grid sizes bigger than animal's home range can be relaxed.

Thank you Shah Nawaz Jelil

Dear Ganga,

Last year I used data from Shannon Kachel (URL from data are in the text) to teach my students how to use SCR in a Bayesian approach, Attached you can find the html with explanations (sorry, in Spanish!). About your question, if you want to use SCR, I think between 7-10 km is fine.

Best regards,

Jose

Dear Gangi

In terms of grid size selection and method of data collection, you may find our recent paper useful. We compared two different methods of data collection for leopards in an Asian montane context, which can be applied for other low density elusive predators in Asian highlands:

Article Citizen science data facilitate monitoring of rare large car...

Darrly MacKenzie recently produced a short video related exactly to your question https://youtu.be/6TtVXy91AY4

Probably less than mean size of snow leopard home ranges in hamalaya.

Shah Nawaz Jelil has it correct, but I'd add that for habitat use you would want the grid cells to be smaller than home ranges.

Ganga Ram Regmi, an old question, but it popped up on my feed.

Defining your grid relative to average home range size for an occupancy study is one option, but it doesn't have to be. It's something that has crept into the literature (particularly for big cats) and stems from a desire to interpret the level of occupancy as proportional to abundance. There's no theoretical reason why you have to do this.

You can define your grid cell size however you want, and I often think it's actually better to define your cell to be smaller than a home range size, as I describe in the video Jeffrey J. Thompson mentioned (thanks for the plug;-) ).

If resources allow using camera traps, then in the first instance I'd be considering using SCR for snow leopards, in which case you want multiple cameras per home range to get information about spatial movemement.

Hope that helps.

Hi Darryl I MacKenzie, just curious why having smaller grids doesn't breach the 'closure' assumption? Doesn't the species need to be present at a site (grid) for the duration of the sampling season? Im assuming big cats would move between smaller grids on a daily basis.

Good question Peter James Mcdonald. Might violate the strict closure assumption, in which case you interpret 'occupancy' as 'use' provided any change are random at the timescale of your repeat surveys.

Some may not be happy with that, but you'll actually end up with a better representation of the species distribution because with large cells you end up calling the whole area occupied or used, even if the species is actually only present in a portion of that large cell. Bit like how you get a clearer image on a tv with more, smaller, pixels.

Now with camera traps, if you're 100% upfront about it, you're only sampling a really small area irrespective of the grid spacing, then you make an assumption about what area around the camera is your 'site'. The larger you make the area, the more of an assumption you're making, in my opinion.

Thanks Darryl

Hi dear in china we are mostly using 4x4 km2 grid size for north china leoaprd,but its depends on your area,and camera availability,mostly we installed two or four camera in each gride ,three camera for photography and two used for vediuos capture.if you want prences only then need signs suryes first ,otherwise need to indentify humance activite sites,road accessibility,

Its important to mention the objective of the survey before jumping on to the grid size discussion. The size of the grid might vary whether you want to just know the presence/absence, occupancy or abundance. In case of occupancy it was presumed that the grid size should equal the mean home range of a species but smaller grid sizes give better representation as Darryl I MacKenzie explained very nicely. Some even believe that regular close random points give even better representation.

It totally depends on the total area you want to cover. Generally we use 5*5 km2 for a large area. Even 3*3 km2 is good enough for moderate to small area.

dear if you need more scientific method.then kindly check out the home range for both sex or territory markings range.then design your grids size.because transect not possible for such a rugged terrain.