That's kinda like saying your ethernet cable needs an IP address. It makes little to no sense.
A virtual LAN is a way for your switchgear to handle multiple broadcast domains and control frame ingress and egress. Internally, virtually every modern switch silicon supports the concept of vlan's, except perhaps the cheapest of the cheap cheap.
The terminology vendors use can sometimes be atrocious and it isn't always clear what's meant. "Native VLAN?" There's no such thing (and I'm calling it out because it's an oft-used stupid term that is confusing, it's like calling your residential NAT gateway a "router").
Your switch internally deals with tagged frames. Tags are the essential thing used to create networks on your switch. Every frame has a tag.
A frame enters the switch. During ingress, one of several things can happen.
1) If the frame lacks a tag, i.e. coming from a device that doesn't have VLAN's configured, one of two things:
1a) the switch MUST add a tag. This could be the default tag, usually "vlan 1". The assigned tag is called the PVID. Each port can have a different PVID.
1b) the switch could also reject the frame, dropping it. This might happen where a port is marked for tagged traffic only.
2) If the frame has a tag, i.e. coming from a VLAN-configured device:
2a) the switch accepts the frame as-is
2b) Or the switch could also reject the frame, dropping it. This might happen where a port isn't allowed access to a vlan.
For egress, you have a simple list of port memberships. Each port can emit frames for VLAN's that it is a member of. This can be either tagged or untagged. Untagged memberships should be limited to one for each port, and should also match the PVID, unless you have special magic sorcery reasons. Which do exist.
Now, your switch may very well hide a lot of this crapola. And it *is* crapola. But it is also how the sausage is made.
There is nothing magic about a VLAN. It is a virtualized LAN. It is nothing more. It requires no IP address. It does require proper configuration to have a chance of working properly. The beautiful thing about VLAN's is that you can sit at your desk and design and implement your network without running new cabling. You can also run a bunch of VLAN's over a single cable.
So on the FreeBSD side, it's much the same.
You do need to do some configuration to cause a VLAN to work. You need an uplink port, let's call it "em0".
For "em0" you want to configure it only as "up".
Then you want to create a vlan on top of it. Or ten. Or a hundred. Doesn't matter too much.
For each vlan you configure, most of the normal ifconfig stuff you could do (configure IPv4, v6, bridges, etc). is available to you.
You can absolutely create a vlan that is simply marked "up" and then attach a bridge to it. This sounds like what you're trying to do. I've done it, for years, works fine. The FreeBSD vlan* interface is created on the fly and is a virtual (as opposed to a physical) device that performs communications on the physical parent device.
The question is whether it will get configured correctly if you do it via the FreeNAS GUI. That, I don't know.
