An interesting web article/ blog entry by Matt Blaze (2009), sharing experiences and techniques in field recording:

http://www.crypto.com/audio/soundscapes/

What about crowd sourcing of the recordings... So that analyses are prioritized instead of equipment. Much like SETI's use of idle processing time of computer users. Alternatively use an app to gain access to the superior recording functions of new phones. Since frog calls are confined to a certain set of frequencies.Durability could be guaranteed through cloud storage.

https://www.globalsoundscapes.org/record.php

http://www.theverge.com/2014/8/28/6071399/scientists-are-recording-the-sound-of-the-whole-planet

http://setiathome.ssl.berkeley.edu/

https://boinc.berkeley.edu

I use a similar combination as Ariel suggests: Sennheiser ME66/K6 microphone attached to a Marantz PMD661 solid-state recorder. A bit expensive, but robust enough to be used in tropical climate, easy to use, and high quality recordings. I would certainly recommend it.

We mostly use Sennheiser ME67/K6 microphones attached to a Roland R-26 or a Marantz PMD661 (MKI or MKII - NOT MKIII!) solid-state recorder, set to record at 96kHz and 24-bit. These provide with professional, publication-grade recordings of under challenging field conditions.

A somewhat old but still authoritative overview of wildlife recording gear can be found here: https://www.macaulaylibrary.org/how-to/audio-recording-gear/