Trial by Fire

by Roberto

If per chance you hear the phrase "press dis­tur­bance" you might con­jure up im­ages of chaotic news me­dia. What with all the dis­turb­ing re­ports of fake news go­ing around these days, who could blame you? At least, that was my first re­ac­tion when con­fronted with the term "press dis­tur­bance." But worry not reader, this post will not delve into any form of yel­low jour­nal­ism. It turns out that in ecol­ogy, press dis­tur­bance has a very spe­cific mean­ing (and many of you are prob­a­bly well aware of it al­ready). When­ever any ecosys­tem un­der ob­ser­va­tion un­der­goes a stress­ful change it is cal­led a dis­tur­bance. To dis­tin­guish short-term from long-last­ing dis­tur­bances, the terms pulse dis­turbance (short term) and press dis­tur­bance (long term) were coined. In the ini­tial de­f­i­n­i­tions, as ap­plied to "large things" (as com­pared to the "small things" we con­sider in this blog), pulse distur­bances would em­com­pass such thing as droughts, where the num­ber of mem­bers of the dif­fer­ent species oc­cu­py­ing the ecosys­tem would vary greatly but none would be ir­repara­bly lost. In con­trast, press dis­tur­bances would in­volve a pro­longed stres­sor that so changes the ecosys­tem as to change it per­ma­nently. But (and in par­tic­u­lar as they ap­ply to mi­cro­bial ecosys­tems) the terms are now used to re­fer sim­ply to short term ver­sus long term en­vi­ron­men­tal stres­sors, rel­a­tive to the gene­ration times of the com­mu­nity mem­bers. From this per­spec­tive, a press dis­tur­bance would be a stres­sor that, due to its du­ra­tion, has the po­ten­tial to al­ter com­mu­ni­ties be­yond the pos­si­bil­ity of re­cov­ery. It is in­deed a very in­ter­est­ing ques­tion to pose: Will mi­cro­bial com­mu­ni­ties that un­dergo press dis­tur­bances have the re­silience to bounce back to their pre-dis­tur­bance state or will they be al­tered for­ever? One pos­si­bil­ity might be to ask whether a person's gut mi­cro­biome is or is not re­silient to the press dis­tur­bance of months-long an­tibi­otic ther­apy? For an ex­am­ple of how this ques­tion has been ad­dressed see here. But of course, there are myr­iad other press dis­tur­bances on mi­cro­bial com­mu­ni­ties wor­thy of in­ves­ti­ga­tion. With the in­crease of an­thro­pogenic in­flu­ences in di­verse nat­ural set­tings, it is in­ter­est­ing to in­ves­ti­gate the ef­fects on mi­cro­bial com­mu­ni­ties that hu­man-ini­ti­ated press dis­tur­bances have al­ready had. One par­tic­u­lar ex­am­ple that drew my atten­tion re­cently in­volves the some­what bizarre phe­nom­e­non of coal mine fires.

Fig­ure 1. Im­age of a coal fire that has been burn­ing since 1916 in Jharia, In­dia. Source

Fires in un­der­ground coal de­posits oc­cur nat­u­rally, with some of them burn­ing for mil­len­nia. How­ever, the num­ber of coal fires has in­creased dra­mat­i­cally world­wide since the In­dus­trial Rev­o­lu­tion as a re­sult of hu­man acti­vi­ties, pri­mar­ily min­ing. This is be­cause once mines are aban­doned coal seams left ex­posed can eas­ily ig­nite. Worse yet, aban­doned mines are of­ten the sites of land­fills and the trash de­posited there is of­ten burned and quite of­ten this leads to the trig­ger­ing of coal fires. Once a coal mine fire starts it is vir­tu­ally im­pos­si­ble to extin­guish. Some of the fires that were started by hu­man ac­tivity have been burn­ing for cen­turies. The fires con­tinue to rage un­der­ground as they fol­low the very large coal seams. It is es­ti­mated that there are thou­sands of un­con­trolled coal mine fires burn­ing around the world to­day. One of the most in­sid­i­ous ex­am­ples is the fire burn­ing in Jharia, In­dia, which started in 1916 (Fig. 1). Strangely, there are vil­lages on the ar­eas af­fected by the fire, an in­di­ca­tion of the ex­tremes un­der which hu­mans can live. Per­haps the most in­fa­mous coal mine fire burn­ing in the United States is the one in Cen­tralia, Penn­syl­va­nia.

Fig­ure 2. Grafitti on an aban­doned sec­tion of sta­te high­way 61, which crosses the now nearly ab­andoned town of Cen­tralia, PA. Source

The town of Cen­tralia, Penn­syl­va­nia was in­cor­po­rated in 1866 as a re­sult of coal min­ing in the re­gion dur­ing the mid-1800s. Centralia's pop­u­la­tion peaked in 1890 at a lit­tle bit un­der 3000. The pop­u­la­tion de­creased as coal min­ing ac­tiv­ity slowed down but there were still about 2000 peo­ple liv­ing in Cen­tralia in 1962. Around Memo­r­ial Day of that year, in an ef­fort to clean the town dump the trash ly­ing there was burned. Un­for­tu­nately, the dump was lo­cated on the re­mains of an aban­doned strip mine and the ex­posed coal seams caught fire. Ef­forts to con­tain the fire proved in­ef­fec­tive. As a con­se­quence, Cen­tralia is now a ghost town; at last count Cen­tralia had seven re­main­ing res­i­dents in 2013. There is cer­tainly an eerie feel­ing for any­one vis­it­ing this aban­doned town, some­thing that can be ap­pre­ci­ated from the graf­fiti cov­er­ing the aban­doned sec­tion of state high­way 61 (Fig. 2). And the fire still burns and it is pre­dicted to burn for at least an­other cen­tury. While hu­mans have al­most com­pletely aban­doned Cen­tralia, what has hap­pened to its mi­cro­bial in­hab­i­tants?

The mem­bers of the lab­o­ra­tory of Ash­ley Shade at Michi­gan State Uni­ver­sity were in­ter­ested in an­swer­ing pre­cisely that ques­tion. To this end they set out to sam­ple the soils of Cen­tralia to as­sess the ef­fects of the coal mine fire on the com­po­si­tion of mi­cro­bial com­mu­ni­ties therein. First, of course, they had to have in­for­ma­tion on the progress of the fires that burned un­der­ground. They were for­tu­nate in that re­gard be­cause the work had al­ready been largely done for them by Jenni­fer Elick of Susque­hanna Uni­ver­sity and pub­lished in 2011. Us­ing ther­mal in­frared im­agery and com­pil­ing lots of prior data, Jen­nifer had de­ter­mined that there were two fire fronts ad­vanc­ing east to west at a rate of about 2 meters/year. As the fire burned un­der­ground, the soil tempera­tures at the sur­face in­creased, some­times reach­ing 80˚C. But, as the fire ex­panded into new ar­eas, it also re­treated from some of the pre­vi­ously af­fected sites. Thus, the ad­vanc­ing fire pro­vi­ded a per­fect sce­nario of press dis­tur­bance, where some ar­eas were af­fected by the higher tem­pe­ratures due to the fire for sev­eral years but were then re­turned to their prior tem­per­a­tures as the fire ex­panded to new ar­eas.

Fig­ure 3. Soil sam­pling sites at Cen­tralia mine fi­re. In to­tal, 18 sur­face soil sam­ples (5.08 x 20 cm PVC core) were col­lected along two fire fronts in Cen­tralia, on 15/16 Oc­to­ber 2014. Sam­pling si­tes en­com­pass a gra­di­ent of his­tor­i­cal fire acti­vity (red flags: Fire-af­fected in 2014 (tempera­ture > 21°C); yel­low flags: re­cov­ered in tempe­ra­ture, post-fire; and green flags: ref­er­ence soils). Source

Now, the changes in the mi­cro­bial com­mu­nity com­po­si­tion could have been fol­lowed along time at in­di­vid­ual sites in a lon­gi­tu­di­nal study. But that, of course, would have re­quired sur­veil­lance and sam­pling over many years. The al­ter­na­tive was to carry out a crossec­tional analy­sis, sam­pling sur­face soil at many sites along the fire trails such that some sites were un­af­facted (refer­ence), some had been af­fected but were back to nor­mal tem­per­a­tures at the time of sam­pling (re­cov­ered), while oth­ers were still in the midst of years of be­ing fire-af­fec­ted. That is ex­actly what Ash­ley Shade and co-work­ers did. Dur­ing Oc­to­ber 5th and 6th of 2014, they col­lected soil sam­ples along the two west­ward mov­ing fire fronts (Fig. 3). In to­tal they col­lected soil sam­ples from two un­af­fected sites, seven re­cov­ered sites and nine fire-af­fected sites. They took the sam­ples to the lab and pro­ceeded to carry out ex­ten­sive mi­cro­bial com­mu­nity com­po­si­tion analy­ses.

Fig­ure 4. Within-sam­ple (al­pha) di­ver­sity of fire-af­fected, re­cov­ered, and ref­er­ence soils in Cen­tralia for bac­te­r­ial and ar­chaeal com­mu­nity (a) phy­lo­ge­netic di­ver­sity (all P<0.001); (b) rich­ness (to­tal no. ob­served OTUs clus­tered at 97% se­quence iden­tity, all P<0.001); and © even­ness (all P not sig­nif­i­cant). Source

They an­a­lyzed the mi­cro­bial con­tent of their sam­ples us­ing two ex­per­i­men­tal ap­proaches. For one, they cal­cu­lated the num­ber of cells per gram of soil by sepa­ra­ting the cells from the soil, fol­lowed by stain­ing and mi­cro­sco­pe imag­ing. From this they con­cluded that here were no sta­tis­ti­cally rel­e­vant dif­fer­ences in cell num­bers across the three types of sam­ples. For their sec­ond ap­proach they used the sine qua non of mod­ern day mi­cro­bial eco­logy: di­rect nu­cleic acid ex­trac­tion from the sam­ples fol­lowed by am­pli­fi­ca­tion and se­quenc­ing. In their case, they lim­ited the analy­ses to bac­te­r­ial and ar­chaeal small sub­unit (16S) rRNA se­quences. Through quan­ti­ta­tive PCR they de­ter­mined that the num­bers of 16S rRNA copies per gram of dry soil show­ed no sta­tis­ti­cally sig­nif­i­cant dif­fer­ences across all sam­ples, con­sis­tent with their cell count re­sults. From this the au­thors con­cluded that the "over­all communi­ty size is rel­a­tively sta­ble across the fi­re gra­di­ent and that any changes in com­mu­nity struc­ture along the fire gra­di­ent are due to chan­ges in mem­ber abun­dan­ces rather than to dif­fer­ences in the to­tal num­ber of in­di­vid­u­als (commu­ni­ty size) among soils." So, how dif­fer­ent were the abun­dan­ces of dif­fer­ent mem­bers of the samp­led mi­cro­bial com­mu­ni­ties? As per­haps might be expect­ed be­cause of the se­lec­tion im­posed by high tem­per­a­tures for pro­longed pe­ri­ods, the fire-af­fected soils showed the low­est within-sam­ple di­ver­sity, the low­est species rich­ness and were the most un­even with re­spect to the dis­tri­b­u­tions of the species present (Fig. 4). Clearly, the high tem­per­a­tures had dri­ven down the mi­cro­bial diver­si­ty in fire-af­fected sam­ples.
 

Fig­ure 5. Prin­ci­pal co­or­di­nate analy­sis (PCoA) ba­sed on weighted UniFrac dis­tances of phyloge­netic bac­te­r­ial and ar­chaeal com­mu­nity struc­t­ure. Col­ors show the fire clas­si­fi­ca­tion of the soil as fire-af­fected (red), re­cov­ered (yel­low), or ref­er­ence (green). The strength of sta­tis­ti­cally sig­nif­i­cant (P<0.10) ex­plana­tory vari­ables are shown with solid ar­rows. Source

How sim­i­lar or dis­sim­i­lar were the com­mu­ni­ties of each of the sam­ples? To as­sess the ex­tent that each communi­ty re­sem­bled the oth­ers, the au­thors ap­plied prin­ci­pal co­ordinate analy­sis (PCoA) to their data. Sim­ply put, PCoA is a method to vi­su­al­ize the level of sim­i­lar­ity of in­di­vid­ual cases of a mul­ti­vari­ate dataset. Im­por­tantly, this visuali­za­tion method makes ap­par­ent which vari­able ac­counts for most of the dis­sim­i­lar­ity among in­di­vid­ual cases. The re­sults of PCoA analy­ses of the Cen­tralia soil sam­ples showed two re­mark­able facts about the dif­fer­ent sam­ples (Fig. 5). First, the fire-af­fected sam­ples (red dots) were the most dis­sim­i­lar. Not only were these sam­ples un­like the ref­er­ence and re­cov­ered sam­ples, they were ex­tremely dis­sim­i­lar from each other. The high tem­per­a­ture (which proved to be the vari­able with the most ex­plana­tory va­lue) led to ex­tremely dif­fer­ent mi­cro­bial com­mu­ni­ties. The au­thors sug­gest that these ex­tremely dis­sim­i­lar out­comes could be the re­sult of very dif­fer­ent dor­mant (that is, non-grow­ing) species present in low abun­dance in soils prior to the fire reach­ing them but which served as "seeds" that grow and take over once the fire af­fects the soil. Sec­ond, and per­haps most remarkab­ly, the re­cov­ered soil sam­ples (yel­low dots) are quite sim­i­lar to each other and – most im­por­tantly – are sim­i­lar to the ref­er­ence soil sam­ples (green dots). The con­clu­sion: af­ter fire af­fects the soil com­mu­ni­ties, send­ing them to quite di­ver­gent ex­tremes, the com­mu­ni­ties re­cover and re­turn to be­ing close to their orig­i­nal state once the fire sub­sides. In short, soil mi­cro­bial com­mu­ni­ties in Cen­tralia, in con­trast to its hu­man com­mu­nity mem­bers, showed ro­bust re­silience af­ter this press dis­turbance. Con­sider this then: If at times you feel over­whelmed by the con­se­quences of the press dis­tur­bance that the an­thro­pocene as a whole is sig­ni­fy­ing to our planet, you might find so­lace in the thought that once the im­pact of hu­man ac­tiv­i­ties de­creases, the planet's mi­cro­bial in­hab­i­tants will have the re­silience to re­pair the dam­age we have done.

 

Other Posts

  • CRISPR-Based Metabar­cod­ing

    Note­wor­thy — Mi­crobes are abun­dant and very di­verse. But ac­knowl­edg­ing their sig­nif­i­cance in our bios­phere has taken time and re­lied on cu­rios­ity, hard work, and the de­vel­op­ment of tech­nolo­gies to ac­cess, iden­tify, and study mi­cro­bial com­mu­ni­ties.

  • A Mi­cro­bial Utopia: Cheese

    by Nathan Jayne, Amanda Alker, and Nathalie Del­herbe — Over the past decade, there have been many large scale at­tempts to char­ac­ter­ize bac­te­r­ial com­mu­ni­ties, from soil and ocean mi­cro­bio­mes to hu­man mi­cro­bio­mes. There is even a col­lec­tive ef­fort to pro­file the Earth's mi­cro­biome. Sci­en­tists are not only in­ter­ested in know­ing who is there and what they are do­ing, but also how...

  • A Nu­clear Fam­ily

    by Merry — Cer­tain mus­sels called "ba­thy­modi­olins" are part of the spell­bind­ing fauna of the dark world of oceanic hy­drother­mal vents and cold seeps. Sim­i­lar to other meta­zoans in that realm, they rely on chemosyn­thetic bac­te­ria...

  • The Bac­te­r­ial Con­sor­tium Chlorochro­matium ag­gre­ga­tum

    by Mark O. Mar­tin — With apolo­gies to the old Turtle's song (which knocked the Beatle's "Penny Lane" off the Bill­board charts in the spring of 1967), there are many ways that mi­crobes in­ter­act with one an­other. Some in­ter­mi­cro­bial in­ter­ac­tions can be neg­a­tive, rang­ing from com­pe­ti­tion for re­sources to pre­da­tion. Oth­ers can be es­sen­tially in­dif­fer­ent...

  • Life Af­ter CPR

    by Me­chas Zam­brano and Roberto — For most Eng­lish speak­ers, CPR would be as­sumed to stand for car­diopul­monary re­sus­ci­ta­tion, of­ten­times a real life saver. But for many read­ers of this blog, the acronym has a whole dif­fer­ent mean­ing. For the last few years, CPR, has stood for the "Can­di­date Phyla Ra­di­a­tion." The dis­cov­ery of the CPR su­per­phy­lum was (and re­mains) one of the most re­mark­able de­vel­op­ments in our grow­ing re­al­iza­tion that the ge­netic di­ver­sity...