The roots of ecology rise from questions about trophy. Darwin, Elton, Hutchinson...as far back as you wish to trace ecological thought, trophic questions have been central in the discipline. One organism consuming another (or part of another) is critical for population regulation, community assembly, ecosystem-level transfers of energy and material, no less the evolution of species.
Ecologists have been watching organisms consume other organisms for a long time, but, curiously, we still know little about what organisms eat.
Part of the reason is that it is hard to quantify consumption. Visual observations are difficult to translate to data. I've watched bison graze grasslands at a distance of feet and even after 30 seconds couldn't come up with a list of what they just consumed. Worse is that most consumption happens out of view. Dissections or regurgitations are necessarily disruptive and identification is still difficult. Isotope analysis gives crude answers. Microhistological analysis of fecal material suffers from differential digestion and lack of specificity.
In short, food web diagrams are hard to generate.
The rise of next generation sequencing has opened new opportunities to quantify what animals eat with resolution that never existed before. Fecal samples from an animal can be sequenced to determine what plants, invertebrates, fungus, or vertebrates they have been eating.
Noah Fierer and I have been working on helping people understand what animals eat for a bit.
So far, we've sequenced fecal material from over 20 different species that range from bats to bison, from prairie chickens to whooping cranes, from moose to mule deer.
I've talked with each of the ecologists afterwards about their data and there is one commonality to all of these discussions: surprises. There are always items in the diet that the ecologists didn't expect. And not just rare diet items. Perceptions (mine included) were always a good ways off of what the animals consumed most. For example, bison were long considered to eat mostly grass. Not so. Their diet can be dominated by forbs and shrubs at certain times of year.
A good technique is only as valuable as the importance of the question that it helps you answer.
Next generation sequencing of fecals is an amazing technique that will open up major insights into how populations are regulated, how communities assemble, and how energy and materials flow through the ecosystem. The technique will open up new insights into the coevolution of predator and prey.
A few bottlenecks still restrict advances.
1) We need better genetic barcode databases. There are still too many gaps for the technique to be commonplace. Many other people are working on this, but we need to sequence key barcodes for all collected taxa. When a sequence appears, we need to be able to compare it to the sequences from known taxa.
2) We need better species specificity. Give or take, current sequencing of diets can get down to ~genus level fairly well, but we need to develop different techniques like hierarchical sequencing to get us to the species level better.
3) We need to be able to quantify the diets of omnivores better. Currently, we can quantify the relative proportion of plants in diets. Or animals in diets. But, knowing the relative proportion of each is difficult. Over the course of a day, a bear (for example) could eat plants, fungi, insects, and vertebrates. But how much of each? Plus, DNA from the diet of prey is present in fecals, but we cannot necessarily tell whether the predator or prey ate a given taxa (or both).
There will always be more technique development necessary, but the biggest bottleneck is still utilization of the technique. Ecologists need to start using it. We're learning a great deal every time we use the technique.