How do I do this correctly with pytest?

gistfile1.py

@pytest.fixture
def error_condition_one(some_fixture):
    ...set up one type of error condition...
    ...for example, an error response with httpretty, or a mock...

@pytest.fixture
def error_condition_two(another_fixture):
    ...set up a different error condition...

@pytest.mark.parametrize('error_condition', [
    error_condition_one,
    error_condition_two
])
def test_handles_error_conditions_uniformly(error_condition, caplog):
    """Any error conditions should return None and log a warning"""
    assert the_function_under_test() is None
    assert 'bad things happened' in caplog.text

The basic idea here is that I want to verify that the_function_under_test behaves in a specific way under a variety of preconditions. I'd like to capture those preconditions as fixtures.