First: To the journals from which I draw:
I would hope that this is not viewed as any sort of
copyright violation. Believe me, anyone who reads
just my analysis without looking at the original...
probably doesn't care enough about the topic to
be a potential customer for you.

I may not always agree with the author's conclusions,
or I may criticize a technique. If you are an author
and you see me say something you think is wrong,
feel free to contact me at: bckirkup@post.harvard.edu
I only have a limited time to read and digest the articles,
and so my critiques have the potential to be flawed.
I'll gladly correct anything that turns out to have been wrong.

ben

Applied and Environmental Microbiology, July 2002, p. 3432-3441, Vol. 68, No. 7
Modeling the Interactions of Lactobacillus curvatus Colonies in Solid Medium: Consequences for Food Quality and Safety
P. K. Malakar, D. E. Martens, W. van Breukelen, R. M. Boom, M. H. Zwietering, and K. van 't Riet

The big issue:
Bacteria can be grown in a number of different kinds of test media, in the laboratory. Some of these are surfaces, some are solid matrixes, and many are liquids (stirred and unstirred). When bacteria which spoil food are tested for their ability to produce toxins or to cause other trouble for people, their growth dynamics and so forth are measured in the test media. Now, the best test media at a certain level is going to be the food item itself. But real laboratory media can be standardized, defined chemically, and used much more extensively.
So it is important to understand how the growth characteristics of spoilage organisms change among the different kinds of media, and the different kinds of foods, and how to relate the two.

This paper approaches the problem by constructing a mathematical model that looks at how growth and acid production by certain bacteria vary with the type of media, type of food, and density of initial population (inoculation).

The organism they model is Lactobacillus, which produces lactic acid in food. The organism can produce more lactic acid then even it can live in, and sours dairy products.

They create models based on differential equations that track the growth of cells and the production of lactic acid within a colony. Diffusion of the lactic acid will eventually acidify the environment around the colony, making it inhospitable even to the cells which would normally produce it. Thus, colony growth is self-limiting, even when nutrients are incompletely utilized.

The authors proceed to create real cultures in liquid and semi-solid media, demonstrating the accuracy of their mathematical model. Despite some deviation at 20+ hours, the model predicted the behavior of the cultures accurately.

The most important result of this is the finding that because of the mass action diffusion of lactic acid, solid media can be modeled by liquid media only when the inoculum size is high, and the colonies themselves begin interacting relatively quickly, creating a somewhat homogenous environment. When heterogeneity is high in the media, that is, when the inoculum is low and therefore spread unevenly on the scale of colonies, the colonies don't interact until they are already self-limited; that is, the heterogeneity becomes accentuated over time and a solid media is needed to model the growth.

This is particularly important if the cells produce, say, some toxin just during one phase of their life cycle.

The results seem rather intuitive, which makes them pleasing though rather unexciting. The models are a bit complex, but easily generalized and constructed in a common sense fashion.

All in all, good stuff!

Applied and Environmental Microbiology, August 2002, p. 3859-3866, Vol. 68, No. 8
Association of Microbial Community Composition and Activity with Lead, Chromium, and Hydrocarbon Contamination
W. Shi, J. Becker, M. Bischoff, R. F. Turco, and A. E. Konopka

So... I wrote this up and blogger ate it. *sigh*

Here goes, again.

The Issue... many sites are contaminated with pollutants such as heavy metals and oil, solvents, etc.
Heavy metals here are represented by Chromium and Lead... Chromium is used in cleaning agents and paints,
lead in paints, solder, plumbing, etc. Oil is represented by "Total Petroleum Hydrocarbons" or TPH.
The question of how bacteria can "mineralize" (i.e. eat) the TPH to clean up a site has been asked many times.
People have tried to develop bacteria, use existing wild bacteria, etc, to eat a variety of solvents like benzene
or toluene, as well as crude oil.
Heavy metals, on the other hand, are often best immobilized by plants, which are then harvested and decontaminated.
They tend to inhibit bacterial action, or so conventional wisdom goes.

The question is: how to deal with a multiply contaminated site? Very few superfund sites have a single chemical to worry them.

The authors take a real site where the bacteria have been living in contaminated soil for a number of decades. There are various parts of the soil containing various amounts of Cr, Pb, and TPH.

Three questions:
1. do the bacteria populations vary depending on the pollution levels?
2. do the bacteria have different metabolic states depending on the pollution levels?
3. will the bacteria respond to introduction of "good food" in contaminate sites?

The answers:
1. Two variables determine 92% of the variation in populations between sites. The main variable is TPH. Where TPH is high, fungi are present and bacteria are relatively scarce.
2. The bacteria do get suppressed by heavy metals. However, they have developed a degree of tolerance not found in strains from uncontaminated sources.
Because there is a sigmoidal curve in the response of amino acid uptake and sugar utilization to heavy metal concentrations, the authors suspect that different parts of the population have different tolerances to the metals. I disagree. It looks to me like sub-population structure would create a step-like pattern with several sigmoid curves superimposed. Instead, we see a single smooth sigmoid, which differs only slightly from sample to sample. Thus, there is one messy threshold, which is standard for toxicity analysis of a single population per sample. This avoids their entire discussion about microhabitat and bioavailability.
A very interesting result is that by adding Lead Nitrate to some samples, they kill the bacteria, while in others they stimulate them. Why should the lead stimulate the bacteria? Well, they eliminate the idea that the nitrate is being stimulatory by adding potassium nitrate. One thing is that this effect only occurs in addition of the chemical to whole soil samples, not to extracted cells. Possibly, the chemical is reacting with something in the soil. Another possibility is that they are not measuring increased biomass, but increased sugar mineralization. Possibly the bacteria in that sample have a way to use sugar to pump out lead ions, and they are churning through the energy to survive the lead or repair damage, etc. An intriguing result none-the-less.
3. Adding alfalfa to the soils provided some promising results. In the most contaminated soils, they started slowly, but ended up getting surprising amounts of carbon mineralization. It looks like the alfalfa was successful in convincing the bacteria to eat the TPH... which is exactly the desired outcome.

So, a good paper, kudos to the authors!