I think it would be very hard to disentangle cause and effect in such a study, at least to generate quantitative results.  How would you measure the degree of all human impact? Could you look at island species and compare populations on matched inhabited and uninhabited islands in terms of fecundity, longevity, etc.?

You are in trouble before you even start -- though this is a cool project.

I would try text mining the literature for articles that study specific organisms and report an environmental shift of some sort. I would then continue to search for articles that confirm/refute this find. Except that it is hard to publish inconclusive or negative results. So your literature search is biased in favor of your hypothesis regardless of whether your hypothesis is true or not.

This project requires a careful definition of adaptation (versus a change), and evolution. So there is research of the effect of the A-bomb tests in the Marshall Islands. The tests are detectable in animals like the sea urchin (http://fishbull.noaa.gov/1014/19ebertf.pdf). Is there adaptation? How does this affect other organisms?

I'd try to find causal relationships of loss of habitat to rates of extinction or decline in species population/richness, vs. those species whose population/richness improved, and then look for what adaptation and/or change in ecosystem composition and structure might favored the more successful species..

This sounds to be a huge project. Actually, human impact is measured according to the receiving organism. It has many types such as trampling, fragmentation, chemical compounds (such as oestrogen analogues, secondary metabolites, biocides). On the other hand the impact on the organism may be measured by different selection measures. For example a chemical compound may reduce the reproductive output of an organism while another chemical may cause reduced digestion of food items. Human impact can trigger the release of both types of selection reduction.

As a result, a meta-analysis (=synthesis of already done research) can be vague.

Moreover, the effect on the selection can be a set of responses. For example, the impact of climate change, if this factor can be considered a type of human impact, can be a causative factor for the geographical translocation of an organism or  host change to another organism or fecundity reduction to a third species.

From my studies in Marathon, Attica, Athens, Greece I drew the conclusion that it is impossible to isolate a type of human impact from other factors such as the vegetation change which is caused by a multitude of other human induced types (=factors). The deposition of soil debris in a vegetation type with a sandy substrate can cause the local extinction of sand inhabiting organisms many of which are able to colonize the new soil debris substrate. So, there is a local redirection of the evolutionary process. But this is an ecological effect not an evolutionary one. It is questionable how the human impact can act in evolutionary time and in the same lines it is unknown how its effect -if any- can be appropriately measured.

There is also genetic pollution, as in the effect of salmon hatcheries on the genetic composition of the salmon.

https://www.sciencedaily.com/releases/2011/12/111219152508.htm

This is not an impossible task. You need to define your variables, even for a meta-analysis. What will you measure? What do you mean by anthropogenic "challenges"? Do you mean anthropogenic influences, impacts, or activities?

Here is an example. Coral reefs are impacted by SCUBA diving and other tourist activities.

http://www.dcbd.nl/sites/www.dcbd.nl/files/documents/Barker%2520Roberts%25202004%2520Biol%2520Cons.pdf

https://www.researchgate.net/profile/Jamie_Tratalos/publication/222705382_Impacts_of_recreational_SCUBA_diving_on_coral_communities_of_the_Caribbean_Island_of_Grand_Cayman/links/0fcfd507526083311d000000.pdf

http://myweb.facstaff.wwu.edu/~shulld/ESCI%20432/Zakai2002.pdf

http://www.jstor.org/stable/4314577

Hope these help give you more ideas.

Article Impacts of recreational SCUBA diving on coral communities of...

Is there a database where I could search specifically for papers about evolutionary change, phenotypic change, population genomics and population dynamics?

Libraries, especially university libraries would have these. Agricola, Cabi, web of science, would be good candidates. However, they have their own search engines, and they are commercial products. I don't know of any free open-access equivalents. I have no idea how open these organizations are about their databases. You could also write your own search engine to surf the web, or use google scholar. I would think the hardest part of this activity would be to use enough key words to weed out most of the stuff you don't want without removing things that you do want. You may just end up with a list of a few thousand articles and then have to weed through those by hand using title and abstract. Also make sure that you include spelling variants. I once was interested in Aphis gossypii. On a whim, I tried Apis gossypii. A simple misspelling, but it changes the organism from an aphid to a honey bee. Well, one of the major databases had a paper on Apis gossypii. While I told them of the problem and they thanked me, the error persists. I think they might have added Aphis to the paper, but did not delete Apis. At any rate, the paper was lost unless one tried searching for the mistake.

GenBank and the NCBI database have articles associated with the genetic submissions. It could take time, but if you get the hang of it, that's one option.