Westerman, P. R.
Group Crop Health, Faculty of Agricultural and Environmental Sciences
University of Rostock, Rostock, Germany
It is known that seed predators (granivores) can consume large proportions of newly produced weed seeds. This way, they can prevent or strongly limit new seed additions to the seed bank, which can help reducing future weed problems.
The identity of the main seed predators varies strongly from region to region. Sometimes, a clear reason can be found for the presence/absence of a specific type of seed predator. For example, harvester ants are almost exclusively found in arid and semi-arid regions where the underground storage of dry seeds makes sense. The presence/absence of some granivorous farmland birds has been correlated with seed availability, in particular during the overwintering of the animals and the raising of chicks. However, often a clear explanation is lacking. In some regions, carabid beetles are the dominant seed predator, in others it is crickets, or granivorous rodents.
Knowing the identity of local seed predators is important for three reasons. Firstly, harvester ants and granivorous rodents, in general, seem to be able to eliminate a larger proportion of weed seeds than do carabid beetles or crickets. In the case that harvester ants and granivorous rodents are absent from a region, one could choose to implement measures that support the less effective seed predator that is present, or measures that enable the return of the more effective seed predator. Secondly, knowing the identity can help explain observed patterns in spatial and temporal variability in seed predation and seed preferences. Species of seed predators differ hugely in mobility, body size, daily and yearly activity patterns, overwintering strategy, habitat requirements, etc. Thirdly, because of these differences, seed predators are bound to respond differently to environmental conditions, management, or landscape complexity. When trying to enhance seed predation, approaches tailored towards a specific type of seed predator are likely to be more effective than blind approaches.
An extra obstacle in dealing with the spatial distribution of seed predators is the absence of historical accounts. It is often unknown whether the observed distribution of a seed predator reflects the natural niche or is human-induced. Farmland birds form an exception. In several EU countries, population trends have been monitored since the 1980s. Species distribution modelling (ecological niche modelling) can help to distinguish between natural and human-induced niches, and can help determining the potential distribution of a species. However, for this technique information is required on the identity, presence/absence or density of a species over large areas. This is where I would like to call upon the international community for assistance.
Simple techniques exist to determine whether seed predation is caused by vertebrates or invertebrates (exclusion cages; presence/absence data). It may even be possible to distinguish between the main types (selective exclusions). More complicated, laborious or expensive techniques are required to determine the identity (e.g., camera trapping) and densities (pitfall traps, nest counts, bird sightings, rodent life traps, footprint traps) of the seed predators involved. Depending on interests, experience and facilities, different protocols can be designed to maximize information output.