Human stressors on watersheds: A Scavenger Hunt

INSTRUCTIONS, & SCHEDULE OF WORK

Parts 1 & 2 – try to complete these before arriving at lab, Week8

Parts 3, 4, & 5 – start work during lab, Week 8; continue as homework Lab period Week 9 – TAs available in Hinds for additionalhelp

FINAL REPORT – due in class, Monday June 1 (start of Week 10)

Part 1. Identify your hometown watershed

Required method:

Go to the EPA (US Environmental Protection Agency) website “Surf your Watershed”– typing that phrase into Google will take you there. Follow the instructions to discoveryour watershed – search on your zipcode, town or county name. This is your ‘hometown’.. The page will display a map of the watershed (or multiplewatersheds) associated with your hometown. If multiple watersheds, choose one to focus on – werecommend choosing the one with the most cultural development (urban area, land use for agriculture,amajor ‘polluting industry’ now or in the past) so that you’ll have lots to work with,historically. Another practical consideration for this lab exercise is to choose the watershed thatcorresponds most closely to a single county (or set of adjacent counties), because most cultural historicaldataare organized by political units of thisscale. Be sure to tell your TA what watershed you have chosen so you do not overlap with others in the lab.

Part 2. Plot the history of (human) population size in yourwatershed

The US Census Bureau is the most reliable and internally consistent sourceof information on population sizes. However, Census Bureau data are compiled for cities,counties, and other political units, whose boundaries will rarely coincide with watershed boundaries.You are thus going to have to decide how to proceed in getting a reasonable estimate ofpopulationtrends in your watershed – use what you learned in Part 1 about your watershed boundariesto decide which political unit (or set of units) to use. We will ask you to describe and justifyyour reasoning. Using an entire county will probably be most efficient, because most economicdataare compiled at thatlevel.

Recommended methods of getting population data:

NOTE: Decade by decade data (“decennial data”, “decadal data”) on population size willbesufficient. You don’t need annual resolution and those data would be very hard tofind anyway – the Census Bureau only censuses the US population every 10years.

Easiest method, if it works: Search Google to see if there’s a wikipedia page aboutyour hometown, county, or major city in your county. These articles usually have asection on ‘population size’ or ‘demographics’, with a table listing population sizeby year, based almost always on standardized data from the US Census Bureau. Theauthor of the wiki-page time-line has usually already done the work of compiling data fromthemultiple Census reports needed to cover local history back to the 1800s and1700s. The wiki page alone is NOT a scholarly source.

Alternate method: Go to the US Census Bureau web site, which organizes data by countyand time-interval. Spreadsheets of population data for each county since the firstavailablecensus are available at These spreadsheets are from a 1996 report, so you will need to updatethemwith data from 2000 and 2010; follow the appropriate link on the above page to dothat. Note that county names are not unique nationwide, so check that the county youarestudying is in the correctstate.

Part 3. Plot two kinds of key economic information on time-lines in order to developmorespecific hypotheses of environmental change linked to humanactivities

Use already-tabulated quantitative data to create graphs displaying historical changesin (a) total acres in farms (agriculture) and (b) annual value of products in manufacturing.TheUniversity of Virginia provides data of these types at boththestate-level and, in many cases, the county-level. On the main UVA page, go directly downto “choose a category” – it will then provide you historical data for all decades available.Totalacres in farms is a measure of rural land – over time, quantities will change withEuropean colonization pressure, economic conditions (depressions), urbanization, andsuburbanization. Manufacturing value might have any number of trajectories, and sums across many kindsof industry – plotting these data will alert you to significant changes in the economic andlikely environmental footprint of your hometown, and can be pursued further in Part4.

**NOTE: This data set is not perfect there are many gaps and inconsistencies. If the unit you chose does not contain what you think is representative data decide on another metric and explain your reasoning when discussing uncertainties and methods.

**NOTE: Manufacturing dollars on this website changes units around 1910. Going from being reported in dollars to being reported so that the numbers will fit within the box (usually a factor of 1000, sometimes more or less). Explain what you believe is happening to the manufacturing products and if you correct for this error please explain how and why when discussing methods and uncertainties.

Part 4. Original search for data on an environmentally significant economicactivity

Most of the historical data provided by the UVA website (Part 3) extend up to 1950 orso, stopping far short of today. In Part 4 you will attempt to bring that environmentallyrelevanteconomic history up to date. Using Google, you will find quantitative data for thedecades between the 1950s and the 2000s or 2010s, focusing either on some measure of ‘land use’(e.g., agriculture, forestry, conversion of open lands to housing & malls) or manufacturing(expansion, diversification, conversion, decline of ‘industry’).

**Read all the way through these instructions for Part 4beforestartingwork!**

Important notes about searching for and selecting data

Getting historical series of quantitative data is much harder than you might think, even with the web.Wedo not want you spending too much time spinning wheels. We will thus define “today” as anytime between theyear2000 and today – a relatively coarse definition that is comparable to the decadal temporal resolution of the datainParts 2 and3.

Also, in many cases quantitative historical data areonlyavailable for a larger area than the study area (eg., trends for the state, or for the US, rather than just for the county or watershed of interest) or are only available for an adjacent area. Such data are used as asubstitute for more specific data – we flag them in our discussion, considering what error this might introduce intoour analysis, just as you identified and discussed gaps in time-series data in Parts 2 and 3. Thus, if you run across atime-series of data for a larger or nearby area during your search for data about your watershed, make a note of it–continue your search for something better, but you can fall back on the spatially coarser or analogous data. Sinceweare interested in getting a historical perspective on stress on a watershed, it would be best to select a substituteareathat drains into the same larger body of water. For example, if your Illinois watershed drains to LakeMichigan,choose if possible a substitute area that also drains into theLake.

Ways to find data on the history of economic activities for this lab include wikipedia pages onyourtown/county (cite the original source of the data in addition to the wiki-page that posts it), professionalgroups(e.g., reports by unions, by companies, by industry promoters, which will have different biases), andagencies(municipal, state, federal annual reports or summaries).

  • Agriculture/Forestry: Searching Google forsomethinglike“history of dairy production” comes up with a remarkable number of leads to follow (~10 organizations)…“historyof xxx” is, in general, a useful search string. If you have an agricultural school in your state (usually a‘land-grant’university), it will probably have a ‘cooperative extension’ that provides advice to farmers; try combiningthatphrase with ‘dairy’, ‘chickens’, ‘corn’, etc.
  • Manufacturing: wiki-pages for counties etc can beboosterish,downplaying a history or continued existence of polluting industry. Try public library web-pages for morebalancedinformation, and for links to local historicalsocieties.

Citing data sources: As a citizen and a scholar, remember that data may be ‘free’ in the sense that theyarenow widely shared, but data still have a source – an originator of the content – whom must be identifiedandacknowledged for their intellectual contribution. You also want to know this source in order to get a sense ofdatareliability – data about whales from acommercialwhaling organization and data from a UN commission may be very different. Content on the webpage of a government agency oraprofessional organization can be assumed to have originated with it, unless noted otherwise (e.g., the page citesthedata or figure as having come from some other source). Thus citing the page – or .gov – willbesufficient for this lab. Wiki-pages are unauthored, but you must still cite the page ( and the source thatisreferenced for the data you are using (these sources are usually numbered, with the full citation provided at theendof the wiki-article). If you find data on the web that you want to use but the source is unclear, then cite thepage( and add “original source uncertain”; treat these data skeptically until you find corroboratingevidencefrom a certainsource.

Finally, remember that you are searching for quantitative data that can be plotted. Data might beprovidedin the form of a table, like the data used in Parts 2 and 3, but data (useful numbers) might be sprinkled throughatext. Don’t be distressed if the data have many gaps – gaps are endemic to historical information. Inprofessionalwork we pound away to acquire gap-free and internally consistent data, but at the end of the day we must still,likeyou, describe and consider incompleteness in weighing the confidence of ourinterpretation.

Part 5. Develop testable hypotheses of environmental change: Predict the consequencesof stressors & theirinteractions

In this last section, we want you to think about the effects of individual stressors and of interactionsamongmultiple stressors, creating testable hypotheses (predictions) about environmentalchange.

Cause & effect (Stress & response): Your four plots of potential or suspected stressors on thenaturalenvironment – human population size (Part 2), area farmed, manufacturing output (Part 3), and a morespecificaspect of land use or industry (Part 4) – constitute a series of hypotheses for environmental change in thewatershedand downstream water bodies. Your initial prediction would be that the natural environment changed,andspecifically deteriorated, in parallel with change in the stressor – deterioration increased when the population(oreconomic activity) increased, leveled when it leveled, and declined when it declined. Depending on the kindofhuman stressor, the environmental impact might have been on water quality, quantity of natural habitatandbiodiversity, promotion of aesthetic or recreational species, etc. all of these, and others, are possible responses (see Figs. 1, 2, & 8forinspiration).

Multiple stressors: Environmental response to stress can be a function of timing – if a new stresscoincideswith the application of another stress or arises when other stresses already exist, then the response might bestronger(more negative) than if it occurred alone. On the other hand, it is possiblethatone stressor counteracts another –for example, invasion of a water body by a wildly successful non-nativefilter-feeding clam, capable of clearing vast amounts of phytoplankton from the water column, would counteractnutrientrunoff from surrounding land, which would otherwise drive phytoplankton populations up, increasing water turbidity and suppressing native seagrass communities. This is arguably what the Asian mussel M. senhousiaandtheZebra mussels are doing today in San Francisco Bay and the Great Lakes, respectively (although probablyatconsiderable cost to the diversity of native filter-feedingorganisms).

Extending time-series to the present day: To create a net-stressors curve for the entire 20thcenturyandearly 21styou will need to extend the agriculture and manufacturing curves of Part 3 past the 1950-limit on theUVA website. Unlike gaps in data where you have values on either side, permitting you to interpolate valuesinbetween, in this case you only have a value on one side of the gap. Extrapolating trends beyond available data isaperilous business! And so, if in Part 4 you were able to find the same kind of agricultural or industrial datafrom1950-2010 as the U VA website provided for earlier periods, please proceed with the instruction for building anet-stressors curve. However, for the other category of U VA data, we recommend that you assume (in the absenceofother evidence) that your local manufacturing and/or agricultural trend follows national ones. You can usetheattached graphs (from the Bureau of Labor Statistics and the USDA Census of Agriculture) to fill inmanufacturingand agricultural data for the second half of the 20thcentury.

NOTE: If the data you located for Part 4 contradicts the national trend (for instance, the number ofacresunder cultivation for 1950-2010 increases over time, rather than decreases as is observed nationally), you maychooseto modify your net-stressor curve appropriately. Make it clear in your answer that you have done this.Realize,however, that the fates of particular industries may not be representative of total manufacturing activity inyourregion. And so if your own data are quite specific (eg., dairy industry output, number of workers inmeat-packing),it's probably better to stick with the national trend for computing the net-stressorscurve.