I wrote an earlier post on the River Continuum Concept - an idea central to stream ecology and understanding the physical and biotic nature of streams. In short, the RCC sets up a physical predictability along a river's course which then is instrumental in the predictability of biological communities along the continuum. Stream temperatures are thought to follow a similar predictable gradient. They are cold in the headwaters and warm as a river flows downstream. But is it really that simple?

Image from the Creative Commons: https://all-geo.org/highlyallochthonous/2011/06/when-a-tree-falls-in-a-stream-theres-always-something-around-to-make-use-of-it/

Temperature is almost certainly the most important factor in structuring biotic communities in streams. In fact, probably the first way we categorize streams is by their temperature regime, particularly their summer maximum temperatures. Streams that get too warm will no longer hold trout and transition to species that can withstand warmer water temperatures. And few species can persist in cold water. In fact, as water temperature decreases, so does biodiversity (see post on biotic integrity). Simply put, temperature is an ecological resource (Magnuson et al. 1979).

Characteristics of different sized streams. From US EPA (Poole et al. 2001)

As we often do in science, we develop models to try to better understand the world around us (read the post, "All Models are Wrong, but Some are Useful" to better understand what we mean by models and the types of models). The simplest conceptual model for water temperatures along the river continuum is one of uniform temperatures (panel A below). However, we know that one to be largely a poor model. A second conceptual is a gradient models that says stream temperatures follow the RCC and temperatures increase as we move downstream (panel B below). This is maybe the most standard model to describe stream temperatures along the RCC. Panel C incorporates small-scale heterogeneity. And panel D incorporates that small-scale heterogeneity within the gradient.

Image from: McGuire et al. 2014.

Without getting to into the weeds about semivariograms, the graphs demonstrate the difference in the response variable, temperature in this case, as a function of distance between any two points. In panel A, there is no variability in temperature so all points have the same temperature no matter their distance from one another. In B, it is linear relationship, meaning that the further away two points are, the more they differ. In C, differences are a function of distance at small distances and eventually reaches a point where temperatures differences stabilize. And lastly, in D, small-scale heterogeneity (similar to panel C) is embedded in the temperature of gradient of panel B. This is the model that most intrigues me and I think is generally most applicable to most trout streams.

A comparison of water temperature (degrees Fahrenheit) of the larger, less spring-fed West Fork of the Kickapoo (blue) and smaller Seas Branch (orange) which has a greater proportion of spring flow.

A first basic principle is that smaller volumes of water have the potential to experience more rapid temperature changes. This is why small lakes are the first to freeze and large-volume lakes, the last to freeze. It also explains why streams tend to warm as they move downstream because the evenings do not cool them down as much as they do smaller streams (all things being equal). A second basic principle is that the greater the proportion of the stream (or lake) that comes from groundwater, the more resistant to thermal change the water body will be. In the figure above, the smaller, more spring fed Seas Branch is much more thermally stable than is the nearby West Fork. However, as we continue along the West Fork, the temperature will change due to stream shading, spring inputs, and most noticeably, tributaries like Seas Branch that provide cold water to the West Fork. Back to our models, this is, more or less, model D above.

Sources of Stream Temperature Heterogeneity

There are many factors that create variability within the river's continuum and create areas with reduced or elevate temperatures compared to the temperature we would predict based on their position within the river continuum. This is an incomplete list of features that create variability in temperatures along the river continuum.

Gradient - Moving water helps to keep water from warming and helps dissipate heat. One of the reasons that headwater streams are generally cooler is that they tend to have higher gradients. If you've not done so before, take a water temperature above and below a large riffle sometime and see that effect for yourself.

A high gradient, well shaded reach of a Driftless Trout stream - one of my favorite types of reaches on hot, sunny days.

Stream Width - Building on stream gradient, water that is stagnated warms more quickly than does water with a swifter current velocity. Wide reaches tend to warm the water and narrow, typically higher gradient, reaches tend to cool the water. Narrowing streams and reducing stagnating water explains much of how stream improvements have successfully improved trout carrying capacity.
Direction of Flow - All things being equal, streams and reaches that flow North-South and East-West have somewhat different temperatures because of how much sun energy they can receive. Since our sun largely comes from the south, streams are most effectively shaded by trees to their south. This effect can be particularly important in the winter when snow melt can have a negative effect on stream temperature and thus fishing success.

Features like bluffs can influence streams by shading them, being sources of groundwater inputs, and in controlling the stream's width and gradient.

Riparian Zone - The riparian zone generally defined as the terrestrial area near the stream with the greatest influence on the stream. Most typically we are concerned with land use / land cover and how it provides, or does not provide, shade to the stream. Ben Sellers' graduate project in the West Fork Kickapoo River watershed is insightful about how shading in the "right" places can have a rather drastic effect on stream temperatures not just locally (Sellers 2024).

Brandon snorkeling a sand bed stream of the northern Driftless. These streams are fed by shallow groundwater throughout their lengths and remain cold in the summer.

Groundwater Inputs - Most trout streams are heavily reliant upon groundwater inputs to keep them cold, even the "freestone streams" of the north rely upon groundwater to remain cold. Groundwater inputs can be concentrated - like springs - or more dispersed shallow water inputs as we might see in sandy soils. Brandon Thill, a graduate student I mentored, wrote his thesis on the effects of geology on stream temperatures, habitats, and Brook Trout (Thill 2024).

The view up the Seas Branch watershed - this tributary delivers a lot of cold water to the West Fork Kickapoo River.

Tributaries - Tributaries are generally a source of cold water inputs to trout streams, but at times, tributaries may be responsible for warming of trout streams. Quite often tributary density and size is an important determining factor on the quality of trout waters. I can think of many trout streams and rivers whose quality change over their length mostly due to tributary and/or groundwater inputs.

A summer 2024 image of the Jersey Valley "Lake" near the headwaters of the West Fork of the Kickapoo River which had been responsible for warming the West Fork.

Dams - No feature is likely to have as great of an influence on water temperatures than dams. Dams create warmer water behind them but in reservoirs large enough to thermally stratify, bottom-draw dams often create cold "tailwater" streams such as the San Juan, Big Horn, and Green rivers of the Western US and the Delaware, White, and the tailwaters in Tennessee. Small impoundments on trout streams lack a critical mass of deep cold water and warm the water (Zaidel et al. 2021) and in recent decades have been removed for ecological reasons (Bednarek 2001).

Some of these features have only a small scale influence - like a tree that provides a small area of shade on the stream or a small spring whose cold water is quickly overwhelmed by the volume of the stream. Other features have a great deal of influence on temperature and their effects are felt well downstream. Examples of these are a large open pasture in an otherwise wooded stream corridor, a reach with a much higher or lower gradient than the rest of the stream, or a large tributary or a dam.

The Big Spring that forms a trout stream in Southwestern Wisconsin - the small stream above this spring is quite a bit warmer and much less "trouty".

That is enough for one post, I'll continue this in another post that links water temperatures along the continuum to trout population and size structure.