In a study of bacterioplankton in an oligotrophic lake in northern Wisconsi
n, a community fingerprinting technique, automated ribosomal intergenic spa
cer analysis (ARISA), was used to determine the effect of resources and tro
phic interactions on bacterioplankton diversity. Inorganic nitrogen and pho
sphorus (NP), carbon in the form of glucose (G) or dissolved organic matter
extracted from peat (DOM), and carbon and NP in combination were added to
two types of experimental systems. Ten-liter mesocosms contained all compon
ents of the original aquatic community except for large zooplankton. One-li
ter dilution cultures were prepared so that the effects of grazers and phyt
oplankton were removed. During a S-day incubation, bacterial production sho
wed the greatest response to the carbon plus NP treatment in both experimen
tal systems, but bacterial diversity was strikingly different between them.
In the mesocosms, the number of ARISA-PCR fragments averaged 41 per profil
e, whereas the dilution culture communities were highly reduced in complexi
ty, dominated in most cases by a single PCR fragment. Further analysis of t
he mesocosm data suggested that whereas the NPDOM addition caused the great
est aggregate bacterial growth response, the addition of NP alone caused th
e largest shifts in community composition. These results suggest that the m
easurement of aggregate responses, such as bacterial production, alone in s
tudies of freshwater bacterial communities may mask the effects of resource
s on bacterioplankton.