Zoologists have means of grouping animals based on the fossil record and genetic analysis. In modern phylogenetics and taxonomy, it is preferred to apply names only to clades, that is, to groups of animals defined by common descent from some known or hypothesized ancestor. Another term for a clade is a monophyletic group.
This is all good and sensible, but naturally in ordinary language we have all kinds of group names. “Sea creatures” is a perfectly useful term, but of course it includes animals that are not very closely related at all: molluscs and arthropods and vertebrates and goodness knows what, but not _all_ molluscs and athropods and vertebrates and GKW. “Sea creatures” is thus a paraphyletic group. Some older, formal group names in zoology, such as Reptilia, are also paraphyetic, but it does appear that they have stuck and it is too late to dislodge them from the collective minds of zoologists, despite efforts by hardline cladists.
The order Anura is divided, in English, between frogs and toads. This division is mainly made on the basis of dryness of the skin and affinity to water but there are some odd exceptions. As you might expect, the division is somewhat arbitrary and superficial. The toads are not a clade: there are amphibians we call frogs that are closely related to other amphibians we call toads, but not very closely related to certain other “frogs” (if you see what I mean.) Importantly, he said, drawing closer to his point, some languages do not make this distinction at all. I believe in Greek, all of these are called βατράχια. In Spanish the distinction exists (ranas versus sapos) but the boundaries between the groups are not the same as they are in English!
Sometimes, zoologists use a common name that has been used paraphyletically by general speakers, to denote a monophyletic group, by using the modifier “True”. e.g. members of the family Bufonidae are called “true toads”.
In Australia, we have a number of species of large monitor lizard, that we clump as “goannas” in English. (Goanna derives ultimately from iguana.) The largest kind of these is the perentie: we consider the perentie to be a kind of goanna. In aboriginal languages such as the evocatively named Western Desert Language, the perentie is labelled as a different kind of animal from the goanna.
In Malay and Indonesian, kambing can be either goat, lamb or the adults of small varieties of sheep (though note, there are other words for a large sheep, but not to my knowledge other words for goat). This might seem whacky to an English speaker but it is worthwhile noting that the sets of goats and sheep in English are also a bit jumbled. The sheep most of us are familiar with is the domestic sheep, covered in fluffy wool, but the wild sheep (same species, but non-domesticated) are straight-haired. The Barbary sheep are less closely-related to the domestic sheep (or to the non-domesticated members of the same species) than to the domestic goats.
