You cannot loop over the reader
more than once, not without rewinding the underlying file to the start again.
Don't do that, however; transpose the rows to columns using zip(*reader)
instead:
a0, a1, a2 = zip(*reader)
Demo:
>>> import csv
>>> csv.register_dialect('csvrd', delimiter='\t', quoting=csv.QUOTE_NONE)
>>> data = '''\
... c11\tc21\tc31
... c21\tc22\tc32
... '''
>>> reader = csv.reader(data.splitlines(True), 'csvrd')
>>> a0, a1, a2 = zip(*reader)
>>> a0
('c11', 'c21')
>>> a1
('c21', 'c22')
>>> a2
('c31', 'c32')