Which database is best to upload survey data to?

I have only used Birdata so may be biased. But here is some information which should help:

"What’s the difference between Birdata and eBird – which should I use?

eBird’s main focus is on recreational birding and enables users to track and maintain their own life and site lists as well as to find out what other people have seen at specific sites, both within and outside Australia.

Atlas & Birdata mainly focus on structured surveys, using prescribed methods, such as 2-hectare Searches and Area Searches with data feeding into various conservation programs, such as the State of Australia’s Birds reporting, the Australian Bird Index, IBA monitoring and other BirdLife conservation programs.  These methods provide rigorous data for research and conservation.

Recreational birding is also a vital component of the Atlas & Birdata. We want everyone to enjoy what they are doing, as well as contribute to conservation.

Data collected by both eBird and the Atlas are complimentary.  Both have an important role in bird conservation, e.g. providing information for environmental impact assessments, threatened species listings and recovery programs.

Does eBird and the Atlas & Birdata share data?

The Cornell Lab of Ornithology, which runs eBird, has an open data policy which allows anyone to access their data, other than for commercial purposes.  They are very cooperative and supportive of establishing a cooperative relationship with BirdLife Australia. All data entered through eBird is passed onto the Atlas for incorporating into Birdata."

Birdata Help page 

--------------------------------------------------------------------------------------------------------------

More in-depth information gleaned from fb group discussions:

From Joris Driessen - BLA and Birdata.

"there is a data-sharing agreement between BirdLife and Cornell, that does allow data to flow between Birdata and eBird. In fact, if you're a BirdLife member, your old eBird surveys might already have been brought across into the Birdata portal! (So if you sign up and they aren't there, it might be worth shooting the Birdata coordinator (birdata@birdlife.org.au) an email to check). It's important to remember that differences between the two platforms mean that not all eBird data is suitable to be brought into Birdata (although I think info flowing the other way is fine). The data to be shared does have to be transferred manually though, and a manual transfer hasn't been done for a few years, awaiting a smoother transfer process to be developed"

And again:

"When comparing eBird and Birdata you are actually comparing eBird with one particular aspect of Birdata - atlassing. I.e. what you call 'lists' - eBird is focused on pure listing, without the need to consider what you are doing as part of any sort of overarching monitoring effort. That is not criticism - it is simply what eBird is very good at - listing, showing your own as well as area tallies, % of lists within 20x20km grid cell with a given species. In that regard it is easy and straightforward - and built for that sole purpose. However, despite recent changes in the eBird app, it has historically not recorded effort or method or size of survey area.

Birdata atlassing, originating in the 70s (first atlas) and carried on to late 90s when the 2nd atlas formalised methods (2ha, 500m and 5km surveys) from then onwards, is a completely different philosophy: yes, you can do whatever you like (and you can just list or send in incidentals etcetera) but the core of it is focused on fixed survey sites using pre-defined methods (e.g. cover 2ha sites for 20 minutes X number of times a year) and repeating that effort over time. That sort of data is used for long-term trends of terrestrial birds for example. So, it is checks and balances - eBird has limited to no standardisation of survey method, effort etc, but balances that out to an extent by chucking huge data volumes at the problem of monitoring, whereas Birdata is all about standardisation, needing less data to meet the same end, but in order to do so does require a higher level of commitment from observers. Neither is perfect, neither is better or even the best. Different.

Note that Birdata does many things eBird does not - long-term monitoring of migratory shorebirds, WA Black cockie counts, waterbird monitoring, biennial count surveys for Hooded Plovers and other beach-nesting birds every 2 years (covering half the world population), dedicated Bittern surveys and a whole stack more - and continuing to integrate existing BirdLife projects which are not yet online.

And as mentioned earlier today - if you do like eBird better but still want to contribute to BirdLife monitoring data you can do so by simply going to your browser and marking your lists as 2ha / 500m / 5km surveys. Easy done. Note that we have currently little use for eBird style lists - we can access them, but due to our different philosophies we cannot really use them (at least not for trend purposes)."