Will Durrant poetically describes the diversity that exists on this impartial planet. The most diverse and complex members belong to a group called metazoa in scientific terms; animals in lay. The term animal can be misleading due to culture, but humans and fish, dogs, cats, and humming birds, butterflies and worms all belong the group Metazoa because they are all evolutionarily, although very distantly related.
Phenotypic diversity and differences exist because they are the fabric of natural selection and evolution; the DNA housed in the cells of every living thing produces the phenotypes. However, molecular biologists have thought long and hard about the relationship between complexity and total amount of DNA and have come to the conclusion that the total amount of DNA is not what gives rise to complexity or diversity. Advances over the past decades have cleared up some of the confusion. For example, the genome of the multicellular eukaryotic microbe, Amoeba proteus is 100X the size of ours; humans.
Conversely, many other microbial eukaryote genomes are 100X smaller than ours. Another problem with the the size matters hypothesis is that complex animals such as octopus, humans, and elephants have less streamlined genomes than we would expect; an economical use of the DNA. It turns out that our genes are interspersed with vast DNA deserts that don't code for any proteins as far as we know. These three conclusions are called the C-value paradox (Moore G.P. 1984 "The c-value paradox"): Complexity doesn't coorelate with genome size, organisms that look the same often have different genome sizes, and seemingly complex organisms have more DNA that doesn't make proteins than DNA that does.
If genome size really doesn't matter, complexity must depend on how the DNA is used; how the DNA is regulated. When all of the known modes of gene regulation are combined, some extreme variations in proteins are possible. However, of all known modes of gene regulation, alternative splicing represents an animals number one mode of increasing the number of distinct proteins (Nilsen & Gravely 2010); of producing multiple mRNA transcript from the same gene. 95% of human genes are alternatively spliced, and the single gene Dscam in Drosophila makes over 38,000 products from it's 95 alternatively spliced exons.