Argo Profiles bio-regions labels

[Mayot et al 2026](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2025GB008930) provides a natural description of bio-regions in the Southern Ocean ([supporting data here](https://zenodo.org/records/18526080)).

It would be very nice for end-users to be able to automatically get the bio-region of Argo profiles.

Such new API could look like this...

### From the `ArgoIndex`

```python
from argopy import ArgoIndex

idx = ArgoIndex().load()

# Make predictions for each profiles:
# (internal structures will have a new attribute with bio-regions IDs for each profiles)
idx.labels.mayot2026.predict()

# Export bio-regions IDs in a new columns of the dataframe:
idx.to_dataframe(extra_columns='Mayot2026')

# Search profiles labelled in specific bio-regions:
idx.query.bioregions(labels=3)
idx.query.bioregions(labels=[1, 3])
```

### From the `ArgoFloat`

```python
from argopy import ArgoFloat

af = ArgoFloat(4903970)

# Make predictions for each profiles of this float:

# (internal structures will have a new attribute with bio-regions IDs for each profiles)
af.labels.mayot2026.predict()

# or directly return lat/lon/time/label columns:
df = af.labels.mayot2026.to_dataframe()
```

### From an Argo xarray dataset

```python
from argopy import DataFetcher

f = DataFetcher(ds='bgc-s').region([30, 120, -60, -30, 0, 1000])
# or
f = DataFetcher(ds='bgc-s').float(4903970)

# Get the data
ds = f.to_xarray()

# Make predictions for each profiles:
ds.argo.labels.mayot2026.predict()
ds['labels'] # N_PROF x N_LEVELS or N_POINTS
```

Start with the ArgoIndex, ArgoFloat, and DataFetcher entry points shown in the issue, then read the Mayot et al. 2026 paper and its supporting data. Done means bio-region predictions are available through the proposed labels APIs, can be exported or queried as shown, and work for index, float, and xarray dataset workflows.

Written by the indexing model from the issue text.