A new forecast shows where an infamous aquatic plant might spread next as it moves north.

When weeds are pushed apart, most of them perish. One kind acts in the opposite way. Each shattered piece of it has the potential to float away, establish roots, and develop into a completely new plant. An invasive underwater weed has spread through hundreds of lakes in southern Quebec thanks to this capability.

Its progress has only been halted by the cold. Lakes remain frozen for longer in the north, and summers are too short for the plant to take root. A new forecast indicates which lakes are most susceptible to invasion next and indicates that protection is waning.

A marijuana worth observing

The plant is known as Myriophyllum spicatum, or Eurasian watermilfoil. It has been subtly occupying rivers and lakes for decades, creating dense mats just below the surface.

The issue is those mats. As they decay, they draw oxygen from the water, obstruct sunlight, and push out native vegetation. The plant grows back from the pieces after it is clipped.

Grace Fedirchuk, a biologist at the University of Quebec at Rimouski (UQAR), oversaw a research team. She was curious about the plant's habitat and potential future locations. Her group set out to predict how it would spread.

The north is warming.

This plant was kept out for years by the cold. Northern lakes rarely warmed up sufficiently for the plant to grow since they froze hard and remained cold throughout summer.

That hold is being loosened by climate change. Lakes that used to remain chilly well into summer are now warming earlier in the year and staying warm longer as average temperatures rise. The trend is not exclusive to Quebec.

According to a report on a floating weed, warm-loving water plants elsewhere have been predicted to move poleward as the climate warms, settling in newly pleasant terrain.

Constructing the forecast

The team input a model thousands of data of the plant's presence and absence in order to forecast its spread. Two dozen local measurements, ranging from lake size and surrounding population to temperature and precipitation, were included with every record.

In addition to estimating the probability of invasion in other lakes, the model learns which conditions are conducive to invasion. The model was used by the researchers in about 12,000 watershed sites throughout Quebec.

Six scenarios, each incorporating various warming and population forecasts until 2100, were simulated by the team. Such a prognosis is very dependent on the quality of its data. According to a previous study on the same plant, human access and water temperature were two of the best indicators of where it will manifest.

Local data prevails

Here, the study revealed an unexpected finding. Using data from Quebec, the North American region, and the plant's native European area, they ran the model in three different ways. The local version prevailed when they compared which version made the best predictions. Nor by a tiny bit.

Even though the continental or European versions had much larger datasets, the model based just on Quebec data made considerably more accurate predictions. There is a lesson in the outcome. When a species spreads into new areas, its behaviour may differ from that in areas where it has lived for many years. The model had to learn from the edge.

A motivating factor

One response outperformed the others when the team determined which conditions were most important. Rainfall and water retention were the next strongest indicators, after how hot a lake gets and stays.

The number of people that live near a lake was another human component that stood out. More docks, launches, and boat traffic result from congested shorelines. The cannabis travels between lakes in precisely this manner. All you need is a broken stem.

Fragments root in a new lake, cling to trailers and propellers, and make it through the overland journey. This was observed in a previous model that was strongly related to road access and boater movement.

Charting the spread

The region where the weed could spread under the warmest conditions increases by almost five times, sweeping across Quebec's seas that are currently off-limits to it. The largest advances are concentrated in two areas. Both Western Quebec and a lengthy section of the St. Lawrence River light up as newly vulnerable.

They are not arbitrary locations. They sit at the intersection of intense human activity and rising temperatures. These were the two elements that the model determined were most likely to increase a lake's susceptibility to invasion.

Practical next steps

Prior to this study, no one knew where Eurasian watermilfoil would spread throughout the province as the climate warmed. There is now one, based on the finding that local data are more accurate in predicting a moving plant than wide ones.

This modifies the actions that managers can do. They can monitor the forecast and attempt to prevent the weed before it appears rather than chase it after it does. The least expensive strategy to combat an aquatic invader has always been prevention, which is made possible by a more accurate map. The province can now identify where the weed is most likely to grow next because the winter that once shielded Quebec's northern lakes is melting.