The NTL-LTER study lakes are changing. Sparkling Lake looks pristine, nestled in mixed Great Lakes forest of the northern highlands of Wisconsin. Yet, the underlying ecology of the lake is in flux, signaled by the local extinction of the cisco, a native coldwater fish. In adjacent Trout Lake diverse beds of aquatic vegetation have disappeared, and changes are occurring in the bottom-living invertebrate and fish communities. Crystal Lake, a few kilometers away, is experiencing a population decline of yellow perch, a native coolwater fish; abundance is low enough that their extinction appears imminent. In none of these cases is the cause acid rain, or overfishing, or change in landuse. Rather, the recent dynamics of these three lakes are driven by the invasion of exotic species, the rainbow smelt in Sparking and Crystal Lakes and the rusty crayfish in Trout Lake.
When the North Temperate LTER research began in 1981, smelt were not found in Crystal and Sparkling Lakes, nor had the rusty crayfish filled the shallow waters of Trout Lake. Yet, the LTER program was, in a sense, a preadapted research platform designed to detect, understand, generalize, and publicize the events that were about to happen. In Trout Lake this learning was enhanced by our design of a long-term experiment to track the dispersal of the rusty crayfish around the lake and observe the changes in the structure and function of the lake ecosystem. In none of these lakes were introductions made as a part of the designed research. Instead the research took advantage of natural experiments, driven by human action in the region. A key feature of this human action is the proclivity of humans intentionally or accidentally to move organisms from one lake to another.Two of the lakes, for example, are seepage lakes with no surface water inlets or outlets and in the other, the rusty crayfish was first observed at our sampling stations near a popular boat landing for recreational anglers.
Fundamental ecological questions and human issues are embodied in these studies of biodiversity change. These include the role of disturbance, ecosystem transformation, the resistance and resilience of ecosystems to disturbance, the recovery or restoration of ecosystem structure and function, dispersal and island biogeographic processes, human values of native versus exotic organisms, and policy and management appropriate for local and regional biodiversity issues.
Knowledge of the effects of these exotic species on lake ecosystems and the mechanisms by which they cause the transformation of communities and extinctions of species comes from our studies. Understanding of this system constitutes important new knowledge, but on a relative basis, too much attention is given to these issues and not enough to the mechanisms for regional dispersal. How many times have we seen a documentation of a destructive invasion? Many times! How often have we seen the mechanism that caused the change analyzed, modeled, or exposed? Far less often! How many times have we seen serious analysis of the dispersal mechanisms and a consideration of how interlake dispersal can be reduced? Almost never!
First of all, the changes we have observed are dramatic.
In Sparking Lake after the rainbow smelt invaded, the cisco no longer successfully reproduced. Each year the youngest cisco in the lake was one year older, until in 1998 only one extremely large adult cisco was caught in our sampling. At one time, this species had been a very abundant in the open waters. The cisco produced viable eggs after the smelt arrived and they were in good condition and grew well. The eggs hatched, but virtually none of the young-of-year cisco survived through their first summer of life.
In Crystal Lake after the rainbow smelt arrived, the yellow perch declined in abundance and in body condition —an index of a fish’s growth rate. The young-of-year grew slowly compared to the years before smelt and decreasing numbers survived their first summer. In recent years survived their first summer. In recent years no young of year were caught in late summer and the adult population is composed of only a few large adults. Smelt are super abundant in the open waters as adults and in the inshore waters as larvae.
In Trout Lake, the rusty crayfish has slowly dispersed around the shoreline and is now found around the entire lake. As they dispersed, most rooted aquatic plants sequentially disappeared, two other related species of crayfish declined in abundance, and other changes in bottom living invertebrates and some fishes occurred.
Secondly, the mechanisms of change are determined.
In Sparkling Lake, smelt preyed on young-of-year cisco as the larval cisco moved from their hatching sites along the shoreline, as the surface waters warmed, to the deep cold waters. Predation was so intense as the numbers of smelt increased that virtually none of the cisco survived their first summer. So the mechanism is predation by smelt on the young-of-year cisco.
In Crystal Lake, the invading smelt and the native yellow perch had extensive overlap in what they ate and where they lived. This overlap occurred not only for the adults in open water, but for the young-of-year in inshore areas. Crystal Lake is oligotrophic and little production of food occurs in the upper waters. As the smelt became super abundant, competition for limited food resources caused the greatly reduced growth of larval and adult perch and the lack of survival of young perch through their first summer. So the mechanism is food competition between smelt and perch.
In Trout Lake, the invading rusty crayfish became super abundant. They are omnivores that can eliminate their preferred food while they maintain their large population on alternate prey. First to go were the most vulnerable of the rooted aquatic plants, followed by some of the more vulnerable bottom-living invertebrates, and fish eggs which they also eat. The reason for the decline of the two other related crayfish is complex, but seems to include genetic swamping and competition. The rusty crayfish is the largest species of the three crayfishes and has first access to food and mates and will mate with the other species. Because it is larger it is also less vulnerable to predation losses from fish. So the mechanisms include herbivory, predation, competition, and reproductive interference.
Thirdly, the process of regional dispersal is considered.
A telling model of smelt dispersal among lakes in the Bear River Watershed near the Trout Lake Station suggests that only some of the lakes have a suitable habitat for smelt. Only some of those suitable lakes are connected by surface waters allowing smelt to migrate directly from adjacent lakes. Given observed rates of dispersal, the model simulations suggest that if their only way of invading was through connecting streams, only 25 percent of the inhabitable lakes would be invaded by smelt even after 1000 years. But smelt are appearing in lakes without stream connections. If the observed rate of invasion into isolated lakes (via human transport) is included in the model, then 50% of the lakes are projected to be invaded after only 200 years, 75% after 300 years and almost all after 1000 years.
Crayfish also move through the streams, but their distribution again indicates that human transport is a common means of dispersal as they appear in lakes distant from those with existing populations.
The lesson here is that the many lakes scattered across the Northern Highlands of Wisconsin will have major changes to their structure and function from the invasion of species like the rainbow smelt and the rusty crayfish. Once in the region, connecting streams are transportation corridors. However, many of the lakes are not accessible by surface streams, and many of these lakes could serve as biodiversity reserves
