Shapefile · dBASE IV
What you are seeing — you have a folder of files with the same name and different extensions, and the one holding your actual data will not open in anything you have.
A .dbf is the attribute table of a shapefile — the rows and columns, with no geometry. It is a dBASE IV file, a format older than the web, and you do not need ArcGIS to read it.
Three free options, in the order most people should try them: LibreOffice Calc opens a .dbf directly (File → Open, pick the file) and will save it straight back out as CSV. QGIS opens the whole shapefile — drag the .shp in, right-click the layer, Open Attribute Table, then Export → Save Features As → CSV. ogr2ogr, part of GDAL, does it in one line if you already have it installed.
Two things will look wrong and are not bugs. Your column names are truncated to 10 characters — that is a hard limit in the shapefile specification, so coordinateUncertaintyInMeters became coordinate the moment somebody saved it as a shapefile, and the original name is gone. Accented characters may be mangled — a .dbf carries no reliable declaration of its own character set. If there is a .cpg file beside it, that file names the encoding; if there is not, you are guessing, and the usual answer is that it was written on Windows in a single-byte code page.
.shp, .shx, .dbf, and .prj if it exists. Opening the .dbf alone loses the coordinates and the .prj is the only record of what those coordinates mean..shp in QGIS rather than the .dbf in a spreadsheet, if you have the choice — that way the coordinates come with it..prj before you trust the numbers. If it names anything other than WGS 84, your coordinates are not latitude and longitude yet and will be badly wrong when treated as such.This tool reads spreadsheets, delimited text, JSON and Darwin Core Archives — it has no shapefile reader at all, and sending you here with a .dbf would waste your time. Use one of the tools above to get a CSV. Once you have one, we will map its columns to Darwin Core against their values rather than their names, which is exactly the check a 10-character truncated header defeats.
We audited a file GBIF had already accepted and served. Their validator answers will this load; these had loaded. Figures read from ops/data.json, not typed into this page. We do not issue a rating of any kind — every finding names the rule it applied and where that rule comes from.