Population viability analysis (PVA) is a method used in conservation biology to assess risks for specific species. It is traditionally described as the process of calculating the chance that a population might disappear within a certain number of years. More recently, PVA has been explained as a way to combine ecology and statistics, using details about a species and changes in its environment to predict how healthy a population will be and the risk of extinction. Every PVA is created for a specific population or species, making each one unique. The main purpose of a PVA is to help ensure that a species' population can survive and grow for many years.

Uses

Population viability analysis (PVA) helps predict how likely a population is to become extinct and shows how quickly action is needed to help it survive. It also identifies important times or events in a species' life that should be the focus of recovery efforts. PVA helps understand what causes changes in population size, compares different ways to manage a species, and checks how well current recovery efforts are working. It is often used to create plans for protecting endangered species, compare the advantages and disadvantages of different management choices, and evaluate how losing habitat might affect a species.

History

In the 1970s, Yellowstone National Park was the focus of a serious discussion about how to manage the park’s grizzly bears (Ursus arctos). In 1978, Mark Shaffer created a model to study the grizzly bears. His model used random changes and calculated the chances of the bears becoming extinct and the smallest number needed to survive. This was the first Population Viability Analysis (PVA), which is credited to Shaffer.

PVA became widely used in the United States because federal agencies and scientists needed tools to assess the risk of extinction and the possible results of management choices. These tools were especially important for following the Endangered Species Act of 1973 and the National Forest Management Act of 1976.

In 1986, Gilpin and Soulé expanded the definition of PVA to include factors that influence whether a population can survive, such as genetics. The use of PVA grew quickly in the late 1980s and early 1990s after improvements in personal computers and software made it easier to use.

Examples

The endangered Fender's blue butterfly (Icaricia icarioides) was recently studied to provide more information to the United States Fish and Wildlife Service, which is creating a recovery plan for the species. The study, called a Population Viability Analysis (PVA), found that the butterfly is more likely to become extinct than previously believed. It also identified specific areas where conservation efforts should focus. The PVA showed that because butterfly numbers change a lot from year to year, the population must grow at a much higher rate each year to avoid extinction than is usually needed for other species.

After an outbreak of canine distemper virus, a PVA was done for the critically endangered island fox (Urocyon littoralis) on Santa Catalina Island, California. The island fox population is split into two groups separated by a narrow strip of land called an isthmus. The eastern group is more likely to go extinct than the western group. The PVA aimed to: 1) assess the foxes’ risk of extinction, 2) estimate how sensitive the population is to large disasters, and 3) evaluate recent recovery efforts, such as releasing captive-bred foxes and moving young wild foxes from the west to the east. The results showed that the island fox is still at high risk of extinction and is very vulnerable to disasters that happen more than once every 20 years. The survival of foxes on both sides of the island depends heavily on how many foxes are released or moved each year.

Population Viability Analyses, along with sensitivity analysis, can help scientists determine which factors most strongly affect population growth and survival. For example, a study by Manlik et al. (2016) predicted the future of two bottlenose dolphin populations in Western Australia. It found that reproduction rates had the greatest impact on population growth. One dolphin population was expected to remain stable, while the other was predicted to decline if it remained isolated and reproduction rates stayed low. The difference in survival between the two groups was mainly due to differences in reproduction, not survival rates. The study also showed that changes in reproduction over time had a bigger effect on population growth than changes in survival over time.

Controversy

Debates continue about the best ways to use PVA in conservation biology and whether PVA can accurately predict the risk of extinction.

Many scientists believe that having a lot of field data is important for PVA. They say that to predict extinction risks accurately over T years, data from 5 to 10 times T years might be needed. However, such large datasets are often not available for rare species. Studies suggest that only about 2% of threatened bird species have enough data for PVA. This is a major challenge for PVA when used with endangered species, as its ability to make predictions drops quickly when data is limited. Some researchers, like Ellner et al. (2002), say PVA is not useful in these cases and should be replaced with other methods. Others argue that PVA is still the best tool for estimating extinction risk, especially when using sensitivity model runs.

Even with enough data, PVA might still have large errors in predicting extinction rates. It is impossible to include all possible future events in a PVA, such as habitat changes, disasters, or new diseases. PVA’s usefulness can improve by running models with different assumptions and future time frames. Some scientists prefer to use PVA only to compare the benefits of different management plans, such as evaluating resource management strategies.

The accuracy of PVA has been tested in some studies. For example, one study compared PVA predictions with the actual outcomes of 21 well-studied species. It found that growth rate predictions are accurate if the input data is reliable, but it emphasized the importance of understanding density-dependence (Brook et al. 2000). Another study by McCarthey et al. (2003) showed that PVA predictions are more accurate when based on long-term data. However, the real value of PVA lies in its ability to identify and evaluate potential threats, rather than making long-term, definite predictions (Akçakaya & Sjögren-Gulve 2000).

Future directions

Improvements to PVA may happen soon. These include: 1) developing a clear definition of PVA and rules for checking its quality, and 2) using new discoveries in genetics to improve PVA.