iNat is not really designed or meant to show interactions of these types, but the observation fields do allow enough flexibility that it can be done to some extent. Short of putting fields called “Fruit eaten by (1)” and “Fruit eaten by (2)” etc, and the resulting difficulty in searching based on those fields, the only other way is to just list them, perhaps in the description, and just have the field to indicate that there is “fruit being eaten”
I think a more elegant over-all approach could be to make the observation of the tree, and seperate observations for each species that is eating the fruit. Then just create and use a field that behaves like “Similar observation set” (https://forum.inaturalist.org/t/using-the-field-similar-observation-set-for-linking-observations-of-lepidoptera-when-raising-on/1018), using the the observation number of the tree as the linking field. When you want to see all observations of things that eat fruits, you would pick “Observations with this field”, but if you only wanted to see the observations of the one tree and it’s fruit eaters, you would choose “Observations with this field and value”. You could put the link to that in the description or a comment on the tree observation, to make it easy to get to. There may already be suitable fields set up this way, so it would be worth looking.
As an after-thought, you could make three (for example) observations of the tree, and in each put the field linking to one of the taxa that are doing the eating. iNat guidelines (stress guidelines, moreso than rules) are more toward not doing this, but if it were to meet your requirements and you put a comment in the description to state why you have done so, then I think it would fly… just be prepared to have the argument with who-so-ever on the day decides they are the arbiter of rules for iNat, as I am sure it will happen :)