Revisiting the effects of climate change on salamander body size: the role of natural history collections

Our recent paper, The relationship between climate and adult body size in redback salamanders (Plethodon cinereus), found that salamanders were larger in warmer parts of their range. We also found that that body size increased significantly in places where the climate had become hotter and drier.


Image credit: Brian Gratwicke

Small-bodied lungless salamanders breathe through their skin, and tend to come to forage on the surface in cool, damp conditions associated with spring and fall, which is the best time to find them. They have thrived in the cool, temperate climate of the Appalachian Mountains, making this region a global hotspot for salamander diversity. Because of their preference for cool, damp environments, salamander biologists worry that predictions of warmer climates and more intense rainfall events and longer dry spells in between may be bad news for these distinctive creatures.

Several studies using museum specimens found that salamanders in warmer areas have larger bodies, but one recent study suggested that salamanders were actually shrinking in response to climate change. Subsequent papers have dwelled on the challenges of using museum specimens to draw these types of conclusions, but none re-examined the actual phenomenon of the shrinking salamanders. We designed a new study to revisit the question using museum specimens in a way that accounts for some of the previous limitations.

We selected redback salamanders, which are one of the most abundant vertebrates both by number and biomass in forests in the Eastern United States. One classic study by Thomas Burton and Gene Likens at Hubbard Brook Forest in New Hampshire found densities of about 3,000 salamanders per hectare, mostly redbacks.  This wide-ranging, abundant species is also very well represented in museum collections. About 70,000 redback specimens are held in the Smithsonian’s Museum of Natural History, mostly collected between 1950 and 2000.


A Redback salamander Plethodon cinereus specimen at the Smithsonian Museum of Natural History. Image by Brian Gratwicke

A collection of this size allowed us to pre-emptively select our comparison groups in a way that would eliminate sampling bias, maintain large sample sizes and maximize our power to answer the question.  We corrected our samples for potential sources of sampling bias including seasonal collection bias and potential destructive sampling bias. We found that redback salamander body size actually increased 1.8% in the places that had warmed significantly. Our observations do not really shed any light on whether climate-change is a potential threat to redback salamanders, but there does appear to be a measurable effect on the species.

The nature and culture surrounding natural history collections is changing, and very few redback salamander specimens were lodged after the year 2000, restricting the time period we could analyze. This likely is a product both of the ethics debate surrounding indiscriminate collecting, and the growing popularity of new citizen science tools like iNaturalist which create photographic specimens in publicly accessible databases with critical collection information. We were able to use citizen-science databases like the Maryland Herp Atlas  and iNaturalist to verify that the salamanders are still common in all the counties that have warmed significantly, but body size data were not available. We actively support and participate in these non-destructive efforts, but view them as complimentary to well-curated natural history specimens, rather than a substitute.

Brian Gratwicke is a biologist at the National Zoo’s Smithsonian Conservation Biology Institute in Front Royal, Virginia, USA

Multiple stressors and ecological surprises

The expanding global human population, now about 7.5 billion, is increasing the pressure that we as a species put on the environment.  2016 was the warmest year ever recorded, and temperature records continue to be exceeded. Each year, more natural ecosystems are lost to dam construction, deforestation and urbanisation. Rates of species invasion are increasing, and pollution events continue to pressure native wildlife. Many ecosystems are now threatened simultaneously by these multiple human-caused stressors, yet we still know very little about their combined interactive impacts.

In our paper in Geo (Linking key environmental stressors with the delivery of provisioning ecosystem services in the freshwaters of southern Africa) we review the impacts of multiple stressors on ecosystem services in freshwater ecosystems in southern Africa (e.g. the Okavango Delta; see photo). We chose these systems because freshwaters contribute disproportionately to ecosystem services despite covering less than 1% of the earth’s surface. Freshwater systems are also especially vulnerable to environmental stressors and over exploitation, with water and fish protein growing in importance as commodities, and average species population declines since 1970 estimated at 81% (WWF Living Planet Report, 2016). Communities in southern Africa rely on freshwater ecosystems for critically important provisioning services, such as drinking water and food (e.g. inland fisheries)


The Okavango Delta

We found evidence that water resources for drinking, agriculture, sanitation and power are declining because of both climate and land use change. In some areas, fish production increased because of dam construction or species invasions, but these stressors can have negative impacts elsewhere. Evidence also suggests that stressors can interact to alter one another’s impacts or promote the proliferation of further stressors.

Multiple stressors often cause impacts which are hard to predict because of both complex interactions between the stressors themselves, and interactions within communities (such as those between species in a food web). These unpredictable impacts have been termed ‘ecological surprises’ and global analyses indicate that they are very common (e.g. This creates problems for decision makers when prioritising which stressors to manage or control, especially when it comes to the supply of the goods and services which we rely on from natural ecosystems.

We provide a framework to categorise multiple stressor effects on ecosystem services where they can either be additive (i.e. predictable and the sum of their independent effects) or four different types of non-additive ecological surprises.  For instance, nutrient enrichment in Lake Victoria (because of high nutrient inputs from the surrounding catchment) causes low oxygen levels, killing fish (Photo 2). At the same time the nutrients promote growth of invasive aquatic plants (water hyacinth) causing a successive and synergistic multi-stressor interaction whereby the increase in plant biomass triggers further fish kills in the lake. In addition, the introduction of non-native fish (Nile perch) caused a dramatic decline in native fish biodiversity but boosted the overall fishery catch in the lake, benefiting the surrounding populations (see figure below).



With the growing population, it is becoming difficult to protect biodiversity and rely on our planet’s natural ecosystems for food and water security. Multiple stressors are causing a downward spiral, where our use of ecosystem services threatens the environment and therefore impairs the delivery of these services for future generations.  We need more research into multiple stressors and ecological surprises, and much more needs to be done to reduce the impact that humans have on the environment.

Michelle Jackson is a Researcher at Imperial College London.