The key to the answer is buried in the long name of the MOSFET, which is Metal Oxide Semiconductor Field Effect Transistor.
The oxide layer isolates the gate from the channel (drain-source). For the MOSFET to be controlled, you apply a voltage between gate and source, but because of the isolation, the voltage cannot cause a current to flow from gate to source. Gate currents become significant only when you change the voltage, i.e. when you put more charge into the gate or you withdraw charge from the gate. The resistace being formed between drain and gate depends on the field created by the nearby gate, i.e. the gate voltage. Thus, MOSFETs are essentially voltage-controlled devices, not current controlled devices.
A BJT will allow (and require) current to flow into the base when you apply a voltage between the base and the emitter. This current controls the current from the collector to the emitter. Even if you don't care about the "transistor action", the base-to-emitter part will act like a diode, i.e. it will allow current to flow if biased the right way. One can say that BJTs are current-controlled devices more than they are voltage controlled devices, although you first need a voltage as the cause for the current to flow. Now... Current alone still doesn't explain the power (current * voltage) dissipated by the BJT. Just like with diodes, there is a voltage drop (approx. 0.7 V) across the B-E part of the BJT. Multiply your base current with this "diode drop", and you have a power dissipated by the base current.
This explanation is quite short and by far doesn't cover everything, but it gives you at least some ideas where to start digging further.
However, your question makes sense only when looking at either transistor in the context of a whole circuit. One npn BJT alone can be off with no base and no collector current, so there is no quiescent current. Consider a larger logic circuit in RTL (resistor-transistor logic) vs. CMOS (complementary MOS) logic. With RTL, there will always be some transistors that conduct and some that don't, so some will dissipate power and some won't, even if the circuit "does nothing", i.e. doesn't change its state. CMOS will only require some gate (dis)charge current whenever it changes its state.
Something similar is true for MOSFET-based LDOs (voltage regulators) vs. BJT-based LDOs.
In electronics, when we talk about "environmental conditions" we are not talking about the weather.
Environmental conditions means all of the conditions under which the part must operate that are external to itself. For example, ambient temperature, humidity, mechanical vibration, mechanical shock, liquid immersion, caustic chemical spray, or other factors.
While the weather might affect some conditions like temperature and humidity, if a system isn't specifically made for deployment outdoors, we're more likely to be concerned about conditions generated by our own design efforts, like the choice of whether to include a fan in the enclosure to cool the circuit.
In the case of advantages of BJTs over MOSFETs, they're likely referring to BJTs' typical higher tolerance for ESD events compared to MOSFETs, as mentioned in a recent On Semi application note TND6093/D.
Best Answer
It is very dependent on the circuit. See this reply on details why one would use one over the other:
When is a MOSFET more appropriate as a switch than a BJT?
For discrete transistors there are advantages to each. When we talk about ICs and VLSI systems the MOSFET is the device of choice because of the simpler manufacturing process and the fact that they are easier to miniaturize.