New member this year. Enjoying the community; learning heaps.
Re Annotations - Plant Phenology. The values available to select are No Evidence of Flowering, Flowering, Fruiting and Flower Budding.
If the observed plant does not show any current budding, flowering or fruiting, but has evidence of previous fruiting (eg. an empty pod), do we select Fruiting or not select anything at all?
There isnāt, to my knowledge, a set of standardized definitions for the values under the Plant Phenology annotation. There probably should be.
I think to most people, and in most cases, if the seeds (/ fruits / dispersal units) are no longer physically connected to the parent plant, then the plant is no longer āfruitingā. I would not annotate such an observation with any of the available values.
Welcome to the forum!
I annotate those as fruiting if the new season of flowering/budding hasnāt started yet, because you canāt say thereās no evidence of one, itās there.
Hello! I am actually a PhD student who uses plant phenology data from herbaria and iNaturalist for research purposes. I would say that, although the aim of a project could change what someone classifies as āfruitingā (for example, a project could aim to see how long empty pods last on different species), I would not consider evidence of previous fruiting to qualify as needing the āFruitingā annotation. For my research & research I have seen in the plant phenology community, most people are trying to assess current, active phenological stages in particular species at particular times. So I would agree with ddennism that if seeds/fruits/etc. are no longer physically attached to the plant, the plant is no longer fruiting.
I use that phenology to sort photos - if I had an empty pod, I would click fruiting to see if my pod matched.
The plant phenology annotations are used to create phenology statistics than can be seen on the taxon page, for example:
https://www.inaturalist.org/taxa/55837-Prunus-cerasifera
Click the āPlant Phenologyā tab to see the stats. So you should annotate their current phenology, otherwise the stats would make no sense - you can always see some dry evidence of previous phenology. The stats are supposed to show you when a certain species is flowering, fruiting, etc. Just what researchers (and naturalists) like @lbrensk are interested in.
I agree that itās most useful to annotate as the current phenology. For fruiting, I think both the starting and ending points can be unclear. For the start, I often use the withering of the stamens or petals or colour change. The āflowerā parts may be still visible and coloured but the process has switched to seed development. I suggest that the end of this stage is most usefully recognised by seed dispersal. So, an empty capsule, while certainly evidence of fruiting, would not be considered to be part of the fruiting stage. There are some species that retain the seeds on the plant through much of the winter, leading to a prolonged fruiting period. Two western North America examples are Arctostaphylos uva-ursi and Elaeagnus commutata. These criteria wonāt work perfectly for every species but I think will help standardise the phenology annotations.
Welcome to the forum!
As it was stated in multiple threads, stats on taxon page only show what people have met, not actual timings, so you shouldnāt rely on them anyway. And many fruits stay e.g. on trees, for years, and thereās nothing wrong with annotating that.
I use the Observation field: Plant phenology->most common flowering/fruiting reproductive stage:
and choose the option āSeed dispersedā.
Thank you. I didnāt know this existed. The stages are better than the ones in the Annotation/phenology field. Iām not sure when either of these should be used - Annotation vs Observation. Are the phenology statistics based only on the Annoation values or Observation values as well?
The above is a key takeaway.
I think the above is a useful rule of thumb but there are edge cases. The common milkweed (Asclepias syriaca) holds onto its seeds for a long time, sometimes well into the winter months, long after active plant processes have retreated underground. Is such a plant fruiting?
To increase the value of the data, iNat should publish standard definitions (in the FAQ) for budding, flowering, and fruiting. The boundaries are the most difficult to classify. For example, as a plant makes the transition from flowering to fruiting, itās difficult to know when flowering stops and fruiting starts. Are there general characteristics of flowering or fruiting plants that can be applied in such cases?
I am thinking also of Banksias which hold their fruiting cones for many years until fire stimulates the release of the seeds from the follicles.
Hereās observation fields and annotations explained and compared:
https://www.inaturalist.org/pages/annotations
Basically, annotations are maintained by the iNat staff, while observation fields can be created by anyone. This leads to a lot of duplicated observation fields and this problem only increases with time. Thatās why only annotations are used in stat generation. If you search for āphenologyā in the observation fields, you will see many duplicates:
https://www.inaturalist.org/observation_fields?q=phenology
All Iām saying is that annotations are supposed to document the current phenology. If a fruit is on the tree, then I guess it is āfruitingā as currently defined - āFruit visible and still attached to the plantā (shown as mouse-over tooltip on the website). Are you suggesting that we should not set annotations in a well defined manner because the info they provide is notā¦ well defined? This is a logical fallacy called circular logic. We should set annotations in a well defined manner so that the info they provide is as meaningful as possible. Whether you should rely on them depends on many things like - what your specific goals are, sample size, etc.
The boundaries are definitely difficult to classify because flowering to fruiting is more of a continuum, and some taxa are more challenging than others. It also complicates things because you could have a plant with fruits and flowers present at the same time, with certain flowers being farther along in the continuum than others. In fact, the tomato plants in my yard are in this state currently. 
These standards are not officially applied to iNat, but there is a Plant Phenology Ontology, which has specific definitions for different phenological states. A lot of these definitions were based on botanical definitions and/or practices used by the National Ecological Observatory Network, the National Phenology Network, or other global monitoring networks.
Unfortunately, looking through ontological definitions isnāt super user-friendly if youāre new to ontologies. 
But you can look through definitions and terms in the PPO here:
https://www.ebi.ac.uk/ols/ontologies/ppo
Here is the definition of āfruitā according to the ontology: https://www.ebi.ac.uk/ols/ontologies/ppo/terms?iri=http%3A%2F%2Fpurl.obolibrary.org%2Fobo%2FPO_0009001
It is interesting to note for your milkweed question that the ontology does separate the concepts of fruits and seeds.
I wanted to also put the links to āseedā and āfruit presenceā, but iNat apparently wonāt let me use so many links cause Iām new. Oh well!
We donāt have well defined manner yet, and there could be different leftovers from fruiting, not each one would be good enough to be called fruiting, but empty pod is, though Iād prefer to find one with seeds left.
Also, it works well for separating young trees from mature, so why would a person not mark adult trees as such if they were actually fruiting and you see it clearly?
But these annotations are about seasonality, not about life stage. And the stats can show you when (which part of the year) a species is flowering, fruiting, etc. We donāt annotate plants as young, mature, etc. but rather - in which part of its annual cycle the plant is. It is possible to have fields/annotations about something like life stage but thatās even more vague, varying and complex with plants