https://valohai.com/showcase/marais-elephant/

this article mentions some softwares and may give ideas to the proyect.

hmmm... was reading things and this came along

https://www.iothub.com.au/news/iot-enabled-cows-will-humans-follow-494207

...and I thought, what if this UAV works with other sensors -that create a map of how organisms have behaved -  deployed on the ground. Maybe in some animals or even in trees. 

Definitely an interesting idea. As far as I know, UAVs have already been applied to download via radio the stored data from GPS tracking devices, whenever they came along tagged animals. For animals that aren't territorial, this can be an effiicient approach.

thermal imaging isnt low cost at all

and its a serious challenge in the heat of equatorial environments

Right, thermal sensors have failed in many occasions, particularly in tropical countries, but also e.g. in Mongolia, that's why we're investigating spectral particularities in the NIR range. If interested, you might want to check the work of Terletzky et al. 2012 (http://www.tandfonline.com/doi/abs/10.2747/1548-1603.49.4.597?journalCode=tgrs20)

what if the environment is the same temperature as the animals?

i know there are conservation organisations already using drones but they definitely need better drones

Thanks for your feedback. I totally agree, however, most of the time the issue is money. While many projects about rhino and elephant poaching get significant funds, conservation of their habitats and less prominent, though not less iconic (e.g. giraffe) species is always short of funding. That's why I'm looking into a highly efficient way of data acquisition to push costs as low as possible.