Adventures In Citizen Science

: Posted On April 21, 2017

First of a Series on the Changing Face of Marine and Coastal Research and Implications for the Future

By: Harlan Doliner, Chair of the Maritime Group and member of the Environmental Group at Verrill Dana LLP, President of the Marine & Oceanographic Technology Network (MOTN) and Ariadne Dimoulas, Oceanographic Researcher, Marine Educator, and Science Communicator based in Portland, Maine

Could meaningful marine and coastal research data be collected by the public? Could scientists, having larger amounts of useful data than they could obtain on their own, examine the correlations and produce reports shared with the civilian contributors? This partnership between scientists and groups of people who monitor conditions in their environment using reliable standards, and sometimes inexpensive sensors, could be important in promoting collaboration through many different user groups, a process that has been called “citizen science.”

This series will examine where and how citizen science is being utilized for marine research, its impacts on the work of professional researchers, and its implications for science and society. This article focuses on the important work being performed on behalf of the Massachusetts Bay National Estuary Program at MIT’s AUV Laboratory and NOAA’s Woods Hole Laboratory. While we touch on the critical issues of data standards, sampling methodologies, and big data techniques in this first installment, future articles in this series will examine these important factors in depth.

Citizen science in action (May 2015): Researchers, interns, and volunteers tag horseshoe crabs that are drawn to the Massachusetts Bay side of Duxbury Beach each spring for mating. As the crabs are gathered and tagged, their sex, size, and shell condition are noted. It’s part of an effort to gather data on population fluctuations and the movement of the crabs.

Way back in ancient times (i.e., the 1970s), one of the authors of this piece was an undergraduate employed in an aquatic entomology study of Maryland stream life. Following training in classification and individual specimen gathering and identification, he was sent into the field to find habitat, flip over rocks, gather critters, and bring them back to the laboratory for classification, preservation, and (hand written) data entry. The total amount of specimens and related data gathered was the cumulative effort of the less than a half-dozen people involved in the study. Of course that meant that any and all of the conclusions drawn from the study were also the product of the efforts of those few. Imagine how things could have been different if instead of budgeting a few undergraduates and graduate students to “go forth and gather, taking good notes,” a sizable, interested, and motivated constituency, say fly fishers, were trained to do the same thing. The study’s budget would have been stretched and leveraged, a lot more data would have been obtained, and any conclusions or theories engendered by the data would have arguably been considered far more reliable and of greater interest to the public.

“Results [of citizen science] can have real impact.” Pamela DiBona, executive director of Massachusetts Bays National Estuary Program (MassBays), explains: “Citizen science…is about breaking down barriers.” Ms. DiBona and her team at MassBays are connecting citizens to resources that empower all to contribute their skills to coastal science. “People want to know what is being done with their data. Are decisions being made? If the information isn’t looped back to the volunteer, then they may stop volunteering,” DiBona notes, recognizing that a lot of effort must be put into the development of such endeavors, ensuring data are useful and reproducible. Examples of citizen science focused on the Massachusetts coast are found in Summer Science in New England, Herring Run Program, and MassBays’ Citizen Monitoring Network.

Summer Science in New England was an intertidal monitoring program that ran from 2009-2013 with funding from NOAA to the New England Ocean Science Education Collaborative (NEOSEC). During this time, summer campers, aged 12 to 16, carried out field exploration in the intertidal zone and took measurements according to the NaGISA protocol (see Photo 1). With supervision from informal education staff and with guidance from partner scientists, such as science advisor Tom Trott, participants were able to check their observations and identifications. At least 396 youths attended throughout the four years. The 12 camps exceeded the number of sampling days anticipated under the grant by sampling at 12 different sites around New England each year and logging 58 official sampling days. The teenage participants identified at least 77 different species, including one not previously documented in the Gulf of Maine. A colleague at Marine and Environmental Research Institute (MERI) worked with Dr. Trott to confirm that the participants found Hartlaubella gelatinosa, a colonial hydroid, in Blue Hill, Maine.

Citizens have been helping to document herring on their annual migrations up coastal rivers since at least 2010 as a result of herring’s drastic decline from amazing abundance in colonial times. According to the Herring Alliance, some Atlantic coast river herring runs have declined by 95% or more over the past 20 years. In 2006, the National Marine Fisheries Service designated river herring as a species of concern, with numerous factors associated with population decline, including by-catch, habitat loss and degradation, water pollution, poaching, decreased access to spawning habitat, and natural predators.

Summer science in New England. Campers follow NaGISA protocol to sample along a rocky shore at low tide.

Summer science in New England. Campers follow NaGISA protocol to sample along a rocky shore at low tide.

Adventures In Citizen Science

The North and South Rivers Watershed Association’s (NSRWA) Herring Run Program is one example of a volunteer monitoring program that provides information to fisheries agencies. In 2016, NSRWA recruited and trained 68 citizen scientist volunteers to count river herring migrating from March 28 to May 31. Volunteers spent over 193 hours in 10-minute intervals counting over 9,516 fish in First Herring Brook in Scituate, Third Herring Brook in Norwell and Hanover, Herring Brook in Pembroke, and the South River in Marshfield. The results of these efforts led to ongoing discussions around dam removal and pollution management.

MassBays’ Citizen Monitoring Coordinators’ Network arose from the Federal Clean Water Act directing National Estuary Programs (NEPs) to periodically document environmental trends and conditions. For MassBays, with 1,100 miles of coastline from Salisbury to Provincetown, encompassing three bays and 47 sub-embayments, this represents an undertaking beyond the reach of any individual program. Government-led monitoring programs had traditionally provided MassBays with information about water quality, habitat condition, and species status. Over time, these programs encompassed a narrower set of parameters and geographic range. The result was the insurmountable challenge to collect all the needed information over MassBays’ large geographic area without help. As a solution to this challenge, MassBays turned to citizen monitoring carried out by community-based environmental organizations for help. Now, non-profit, citizen-led efforts are the primary source of current water quality and pathogen data for most of the MassBays region.

An example of on-the-job citizen science started in the mid 1990s when Jim Manning, a young oceanographer at the NOAA laboratory in Woods Hole, was on a large government research vessel hundreds of miles offshore on George’s Bank. During Manning’s nightly midnight watch, he would see many, many ships out fishing and could not help but wonder what it was that attracted all those fishermen. A visit to their dock after the research cruise revealed that the fishermen similarly wondered what the scientists were doing. Recognizing that there were many more fishermen than scientists, with many more boats and hundreds of more moorings, Manning first provided these lobstermen with bottom temperature probes to attach to their traps in 1995.

Skip forward to 2001, with funding from NOAA’s Northeast Cooperative Research Programs, Manning founded the Environmental Monitors on Lobster Traps Program (eMOLT) and has, with the help of over 100 New England lobstermen around the Gulf of Maine and the Southern New England Shelf, collected and archived several million hourly observations of ocean bottom temperatures. Now with more than 15 years of web-served time series at several fixed locations, Manning and others can begin to define seasonal cycles, look at long-term climatic trends, and help numerical modelers assimilate these data into their “hindcast” simulations. Like the citizen science performed in Summer Science in New England, this effort took a lot of behind-the-scenes work, including attendance at many lobstermen association meetings on a regular basis, frequent one-onone interaction with the fishermen at the dock, and consistent feedback to the fishermen in the form of personalized mailings. It was necessary to develop a routine and stick with it so the participants knew what to expect every year and see the results on paper. To this day, at least half of the original participants are still fully engaged in the data collection and science.

Jim Manning and Xavier of Manning’s laboratory outfitting a fishing vessel with temperature probe and telemetric capabilities.

Jim Manning and Xavier of Manning’s laboratory outfitting a fishing vessel with temperature probe and telemetric capabilities.

“The fishermen are now witnessing, with data they collected themselves, an unprecedented warming of their waters on the order of a few degrees Celsius in the past decade. These fishermen understand climate change,” said Manning. New advances in the last few years enables telemetering the data in near-real time, displaying and transmitting bottom temperatures as soon as the gear is hauled aboard. The processed data go immediately to the satellite and are posted on a NOAA server. The raw data are uploaded as soon as the fishermen are within Wi-Fi range, without the fishermen having to push any buttons. Over the next few years, these data will be routinely fed in real-time to local ocean modelers for them to assimilate into their forecast projections.

Driven by truncated budgets and personnel limitations, citizen science is now seeing ever-increasing utilization in coastal and marine research. Leveraging non-scientists to make observations or gather data—provided it is performed consistent with applicable standards and practices—goes beyond saving money to dramatically increase the amount of data and data points, often by orders of magnitude. This, of course, results in what is colloquially referred to as “big data,” which in turn leads to the need for formulating data mining and other data analysis methodologies.

With diverse sets of people collecting varied sets of data, analyzing trends can be problematic. It could be considered difficult to combine different qualities of data collected through volunteers. It takes extensive preparation to ensure standards of data collection are accurate and repeatable. In order for MIT Sea Grant research engineers Mike Defilippo and Michael Sacarny to understand the effects of ocean acidification on ocean dynamics and marine life, they are developing ways to combine diverse sets of data through their New Cost-Effective Coastal Ocean Acidification (COA) Monitoring Paradigm.

“Standards are important, but there is a way to combine such [diverse] data,” Defilippo explains in reference to his lab’s ongoing study of Cost-Effective Monitoring System of Coastal Ocean Acidification. The Northeast coast lacks infrastructure for collecting required COA and Ocean Acidification (OA) data. There are no validated simulation models of biochemical outputs such as pH, partial pressure of carbon dioxide (pCO2), total alkalinity (TA), or dissolved inorganic carbon (DIC) in the Northeast. Installing sensors throughout Boston Harbor, Mass Bay, and Gulf of Maine is impractically expensive. As such, MIT Sea Grant recognized the need to create a “cost-effective and robust COA monitoring platform” based on “theory, measurements, and simulation,” knowing computer simulation can help to both increase the accuracy of the forecasting as well as save expense.

Michael Defilippo and Michael Sacarny of MIT Sea Grant displaying their research submarine Odyssey IV at their new Cambridge, Massachusetts office.

Michael Defilippo and Michael Sacarny of MIT Sea Grant displaying their research submarine Odyssey IV at their new Cambridge, Massachusetts office.

MIT’s Sea Grant proposed combining high-fidelity laboratory data, computer simulations, and other measurements gathered through more affordable equipment, combined with all available measurements in the Boston Harbor and Mass Bay as well as all available monitoring stations. This results in a multi-fidelity Bayesian framework in which all uncertainties are considered. Since its beginning, scientific research has been beset by an ironic dichotomy of standards. Even middle- school students just learning about the Scientific Method are quick to cite the need for repeatability of results by applying the same methodologies. On the other hand, researchers pride themselves on developing their own data gathering methodologies they deem the most suitable to their specific projects, organisms, or geographic locations.

The increasing utilization of citizen science introduces more variables and doubts about data obtained by non-professionals. Efforts at cooperative standardization and coordinated training are on the increase.

Examples to be addressed in upcoming articles include the 2016 MassBays Monitoring Summit (www.mass.gov/eea/docs/mbp/monitoring-network/summary- of-summit.pdf); GEOSS, the Group on Earth Observations (www.earthobservations.org/geoss.php); and Zooiverse (www.zooniverse.org).

Remote Explorer (REx IV)—An autonomous vessel for data acquisition and dissemination deployed in Boston Harbor. Researchers funded under MIT Sea Grant will combine and analyze information from monitoring stations in Mass Bay, Boston Harbor, and their tributaries. High-fidelity data are expensive and difficult to obtian, whereas lower-fidelity data are readily avalilable at a lower cost. Integrating multi-fidelity data with the biogeochemical- augmented coastal model (FVCOM-ERSEM) will result in highly accurate predictions of COA. Photo credit: MIT Sea Grant.

Below is a comprehensive list of available measurements in the Boston Harbor/Mass Bay.

SOCAT (fCO2, pCO2, Salinity, T, Depth)

Location: Mass Bay, Cape Cod

Resource: www.socat.info

MWRA (pH, Salinity, T, Depth, Chl, DO, Nutrients)

Location: Boston Harbor, Boston Rivers

Resource: www.mwra.state.ma.us/harbor/html/wq_data.htm

MWRA (Salinity, T, Depth, Chl, DO, Nutrients, pH)

Location: Mass Bay, Cape Cod

Resource: Direct from MWRA

NOAA Stellwagen Bank National Marine Sanctuary (pH, pCO2, Salinity, T, Depth, DO)

Location: Mass Bay, Cape Cod