I don't quite understand the question! Translocation of species is a major tool in restoration.

Translocating analogue species may help restore lost ecological interactions and ecosystem functions such as seed dispersal and pollination.

Griffiths et al 2011 introduced the giant Aldabra tortoise to restore seed dispersal in the endangered tree Diospyros in Mauritius which was thought to be dispersed by a similar extinct giant tortoise. http://www.sciencedirect.com/science/article/pii/S0960982211003460

Further to Lui. Analogue ungulates are widely used in conservation, restoration and recreation of grazed ecosystems. Domestic cattle breeds in grassland where a native equivalent (auroch; wisent) has been eliminated, mouflon on heathland where sheep farming became uneconomic, przewalski where wild horses have gone from the steppe or riverine grassland and huntable grazers where protected equivalents disappeared in southern African savanna. 

There are many native and endemic Mauritian plants that have co-evolved with tortoises, and without the presence of tortoises as tramplers, grazers, browsers and seed dispersers, many plant species are at a distinct disadvantage.  Tussock grasses have declined on Round Island since the removal of rabbits and there is a  need for grazers to keep down the competing exotic grasses.  Now that we have introduced  Aldabra Tortoises as ecological replacements for the extinct Mauritian Tortoise the exotic gasses are being grazed and controlled and tussock grasses are now regenerating.  Without the use of introduced tortoises as ecological replacements we could loose several species of endemic plants that were part of the grazing climax community that was maintained by grazing tortoises and now extinct endemic geese.

Sorry but it is not clear to me if your question refers to Mauritius only or conservation/translocations in general.

The use of ecological replacements has broad currency and can be applied to any system where ecological roles have been lost.   At this stage we should limit ourselves to using ecological replacements that can be easily controlled or removed if their deleterious impacts out weight the benefits.  

Is there any common criteria to follow prior of using analogous species? 

Ecological replacements have of course to be similar in ecology, and we should also look for species that are taxonomically close although this is not essential.,

Until today, conservation was (by definition) "conservative" meaning focused on historically related reference states. In the near future, such philosophy can be questioned when climatic conditions are changing.

Basically, changing environments (eg due to nitrogen deposition) have already challanged the focus on historical reference states or constructed hypothetical references such as "potential natural vegetation"  - see Chiarucci et al. 2011 in Journal of Vegetation Science doi:10.1111/j.1654-1103.2010.01218.x  

However, concepts on novel ecosystems that would relate to future climatic conditions are not well elaborated, yet. This is also true for the concept of analogous species, because there cannot be real analogues but only functionally close species. My opinion is that this field requires a vivid and rapid discussion but action should be considered carefully in order to avoid unexpected negative impacts.

Translocation of species to new environments can result in problems even if this species' niche is not occupied in the target area or was filled by another, maybe locally extinct species in the past. But just monitoring the loss of species and functionality is not a good strategy, either.

Islands lose species..... Even in continental islands in the Mediterranean many many species have gone and are going - often new ones replacing them as humans spread them rapidly (i.e. native trout replaced by continental trout in tiny streams some small catchments in some islands). However shifting baseline syndrome is at work here - what were these freshwater faunas like before man changed everything? Many think many major Mediterranean islands never had any species of freshwater fish for example. Today many of their streams are so degraded that no one can believe a more natural condition may have existed. We write about this in the attached paper and promote the idea of re-introducing an analogue species/populations of the Cyprus river blenny (Salaria fluviatilis) from a nearby population if the local species/population is actually extinct on this island (which we think it is). Of course extreme caution is called for. But re-introduction and scientific translocations etc are interesting in a conservation science framework on islands. (And if we scientists don't get involved, lay persons will spread speceis indiscriminantly). Good questions...