Most likely an elaboration of that argument goes 'humans in general can engage in contractual agreements, although some are physically or mentally incapable of performance or comprehension, respectively. Apes, monkeys etc. cannot do so at all.' One could refute that by observing that animals do sometimes make simple mutual arrangements, such as favorite sleeping spots or the like, and animals who are cosocial with humans can certainly be trained. I presume a contractualist would argue that such inter-animal arrangements are too informal and not abstractable, and that training simply develops reflexive behavior...although since animal training is typically predicated on the concept of earning a reward for performance, this objection seems very much like an attempt to wriggle out of one's performance obligations, and I'm not sure it's conceptually different from overbroad get-out clauses in insurance or employment contracts that basically allow one party to exploit any economic asymmetry in order to avoid payment.