Reconstructing the spread of invasive plants: taking into account biases associated with herbarium specimens