Kartchner Caverns Microbial Observatory
National Science Foundation
photo of caverns


Non Culture-Based Analysis

Impact of Tourism on Kartchner

Nonculture-Based Analysis

Culture-Based Analysis




An initial research goal is to gain a better understanding of the diversity in the microbial community structure in Kartchner Caverns.   To do this we are investigating the variability in bacterial community structure among different speleothems (inter-speleothem variability) or within the same speleothem (intra-speleothem variability).  We are using a variety of approaches to explore these samples including approaches that are culture-based and those that are nonculture-based. Nonculture-based studies depend on analyzing community DNA that is extracted from the samples. Culture-based studies are dependent on isolating bacteria from the samples

Intra-speleothem Variability

To examine intra-speleothem variability in bacterial community structure, three stalactites were chosen as sampling objects in an undisturbed area of the cave. The stalactites are located close to each other, the distance between them is less than 2.5 m swab (Figure1) samples were taken along the vertical axis (Figure 2) of the stalactites.

Figure 1

Figure 1. (left)

Julie Neilson, taking a swab sample along the vertical axis of a speleothem.

Figure 2

Figure 2. (right)

Five swab samples were obtained along the vertical axis of this speleothem.


Total community genomic DNA was extracted from the different samples.  The 16S rRNA gene was amplified using polymerase chain reaction (PCR)( link to PCR graphic).  The primers used for PCR contained a GC clamp for later analysis using denaturing gradient gel electrophoresis (DGGE) and produce a product that is 392 base pairs in length.  DGGE analysis generates a community profile from each sample. This profile is visualized as bands in a lane on the DGGE gel.  Theoretically, each band represents a single bacterial population.  In practice, some bands can contain more than one population and also, some populations can exhibit more than one band.

The DGGE profiles that we obtain from Kartchner samples are analyzed using software called Quantity One ®   .  This program assigns a vertical location for each band present in the gel.  For each sample or profile, a 0 (absent) or 1 (present) is assigned for each possible vertical location.  A similarity matrix is then generated to see how different or similar profiles are to each other. 

An analysis of DGGE profiles from two stalactites shows that bacterial community structure profiles taken along the length of a stalactite are more similar to each other then to community profiles from different stalactites (see figure 3). To further examine inter-speleothem variability in community structure, we have extended the sampling area to include a total of 9 different stalactites and a flowstone which differ in color, size, structure and location within the sampling room (data coming soon).


Figure 3.

Cave Speleothems (right) A photo showing the relative locations of two of the three speleothems sampled for the intra-variability study. These speleothems (cave formations) were sampled in June 2007.; The finger formation (F) on the left is 110 cm long and reddish brown with a coating of cave dust.The white formation on the right is 39 cm long and milky white with little dust. Not shown is the H formation which is 105 cm in length and reddish brown in color. These samples were analyzed using DGGE (Figure 4).

Figure 3

Figure 4. DGGE Gel (below)

The gel shows the microbial community profiles on the F, W, and H formations.  Lanes 1-5 left: F formation from top to bottom; Lanes 1-4 middle:  W formation from top to bottom; Lanes 1-5 right, H formation from top to bottom.

Figure 4


Figure 5. (below ) Non-Metric Multidimensional Scaling (NMDS) analysis of DGGE gel in Figure 4.

NMDS analysis of bacterial community profiles from the intra-speleothem variability study.  Circles delineate three clusters that are significantly different from each other (p = 0.05, stress factor = 0.049, D = 4 where D = dimension)


Figure 5


Inter-speleothem Variability

To examine how physical distance between speleothems affects the similarity in speleothem bacterial communities, we then examined inter-variability among samples taken from the 10 different formations shown in Figure 6 below using DGGE. 

Figure 6. (below) Map of speleothems sampled for the intra- and inter-variability study. 

The large black circle represents the Kartchner Cave room being studied and the speleothems are represented by circles each with its alphabetical label. Arrows and numbers represent the distances between the various speleothems.  The color of each circle shows the clusters formed using NMDS analysis of DGGE profiles.  For example, speleothem G and J (in yellow) communities were significantly different from the rest.  The blue lines show the suggested water source and flow in the room.

Figure 6


Figure 7. (below)  Non-Metric Multidimensional Scaling (NMDS) analysis of DGGE bacterial community profiles from the inter-speleothem variability study.

Labels correspond to the lanes in the DGGE gel in Figure 2.   Circles delineate four clusters that are significantly different from each other (p = 0.048, stress factor = 0.007, D = 4 where D = dimension). 

Figure 7


NMDS analysis of the 10 speleothems revealed four clusters that are significantly different from each other (p= 0.048) (Figure 7).  Examination of these clusters shows that speleothems A and B grouped together and G and J grouped together.  The map shown in Figure 6, reveals that A and B are located at one end of the room and G and J are located at the other end of the room separated from the other formations by a rock wall.  The remaining speleothems, which are all found in the middle of the room, separated into two clusters, one containing E and H, and the other containing W, D, F, and C.  These results suggest that speleothem bacterial communities are more similar among formations that are spatially closer together. It has been suggested to us by the Kartchner Caverns Science manager (Bob Casavant, PhD) that these groupings may result from different water sources that feed them (Figure 2).  Dr. Casavant pointed out patterns of soda straws along the ceiling that are a general indication of water flow. Denser collections of soda straws (young formations) indicate a greater abundance of water.  Lines of soda straws could be followed down the ceiling to the region where the formations were. D,F, and W were all at the base of the same drip line. This drip line forked about 3-4 meters above these formations and one path followed down to the area where the bacon C was located. Thus it is logical that these four clustered together. E and H were clearly below a separate soda straw drip line. Another large crack in the rock which could be traced all the way up the ceiling (between A/B and the C/D/F/W ) separated the flow to A and B from that going to C. This flow pattern came down by the broken shield above A and B.  There was also a lot of Fe rich limestone above the DWFC cluster. The angle of slant of the rock also indicates the flow path. The flow path is consistent with the clusters we have seen.


Our results suggest that the bacterial communities found on samples taken within a speleothem are more similar to each other than to communities found on other speleothems.  We are very encouraged by this study given that the speleothems are actually quite close to each other (Figure 2), with 115 cm separating H and W, and 90 cm separating W and F.  These data suggest:


updated 7/2009

University of Arizona Header The University of Arizona