I think if you search the internet, you'll see the general trend you refer to is fuzzy at best. There are 3 big reasons for the general trend. One involves the synthesis of elements, so there's a Google point for you. Another involves distribution, and selective observation, spatially. Again, distribution of elements is a good Google point for you. Then, there's another, selectivity to the observation, following another variable, which people tend to forget, often even as they go deeply into science. What is the biggest source of error in any observation?