Assessing ant impacts

Knowing whether or not an invasive ant is having a negative effect on people, agriculture or the environment helps with management decisions. Management is expensive and time consuming. In some cases the ants must reach a certain level of abundance before treatment is effective without causing non-target effects. Also, assessing impacts before and after treatment helps understand if any possible non-target effects have occurred, as well as the effectiveness of treatment. If little fire ants (Wasmannia auropunctata) or red imported fire ants (Solenopsis invicta) are newly arrived, there is probably no need to assess impacts before taking action. Why not? Because detection of these species signals an emergency. Just get help to get them eradicated as soon as possible!  Sometimes the impacts of invasive ants are obvious. Clearly the effects of African big-headed ants (Pheidole megacephala), tropical fire ants (Solenopsis geminata) and yellow crazy ants (Anoplolepis gracilipes) on nesting seabirds and their chicks on Johnston and Kure atolls (see image on the right) and Kaneohe Bay indicate a serious problem. Similarly, little fire ant stings are a painful alarm call to action. But in many cases invasive ants are present in numbers that are too low to have obvious effects like these. That does not mean the effects aren't happening. For example, even at relatively low numbers invasive ants can be making the ant communities around them less diverse i.e. fewer species and lower abundances. While many people wouldn't consider that a problem, it can be an early warning signal that the problem will get worse.

African big-headed ants attacking a bonin petrel chick on Kure Atoll, Hawaii. The webbing on the chick’s feet is especially vulnerable (© Cynthia Vanderlip, Maui Invasive Species Committee (MISC) website)

Interactions on food resources : Sampling ant and insect communities : Surveys of other animals :  Effects on people

Read how to assess impacts by ants on other ants and insects, other animals, and on people.

Observing interactions on food resources can give a broad indication of how an invasive ant might be competing with other species.

A note on yellow crazy ant abundance

Measuring yellow crazy ant abundance using card counts (© Antoine Felden, Pacific Biosecurity)

Yellow crazy ants are found throughout the Pacific often in relatively low numbers. However, they do sometimes have population explosions. Yellow crazy ant abundance is measured by card counts. On Christmas Island, where the ants regularly have these population explosions, Parks Australia North uses a card count threshold of 37 to decide if an area should be treated. If the average card count in an area is over 37, the ants are starting to cause harm to the red crabs there (the key species in the environment).  We recommend this measure as an objective way to decide on management of yellow crazy ants in the Pacific as well, based on experiences in Tokelau.

Note that although yellow crazy ants usually have effects  at high abundance, in rare situations they can cause problems for birds and other animals when they are at low abundance.

Interactions on food resources

Placing lures with a high value food resource in uninvaded and invaded areas can help identify which species invasive ants might be competing with most. 

Lures can include meat / fish products (which attract scavengers such as crabs), fruit (which attracts some land birds), or seeds (which attract rodents such as rats). 

The lures are placed in the environment and observed, either continually, or after a period of time has passed. This can also be used as a first step for identifying potential non-target effects of pesticides, as part of environmental impact assessment.

Back to top

Sampling ant and insect communities

Other ants (and insects) are typically the first affected by invasive ants. Pitfall traps, which are often used to estimate the abundance and diversity (number of species), are also used to assess impact. To determine the severity of the impact, pitfall traps from areas where the invaded ant is present are compared with areas where the ant is not present. If the invasive ants are having an effect the number of species and their numbers will be substantially lower in the invaded areas. The method is quite time consuming, and requires replicates (more than one site).

Pitfall trapping is fine for ants that spend most of their time on the ground, but aren't going to be effective for species like little fire ants which live predominantly in trees. If you find little fire ants (or red imported fire ants) there's really no point in assessing impact - just get help to get them eradicated as soon as possible. 

Pitfall trapping is also great if you can easily identify the species you catch. Sometimes that isn't possible without the help of an expert. But you may be able to count the number of the invasive species in the traps, and sort the rest into groups that look similar. Instead you could do a visual survey of other insects such as spiders and other ants in the invaded area and compare them with the uninvaded area.

Back to top

Surveys of other animals

When invasive ants are at lower abundance seeing actual interactions between them and native species can be difficult, so observing the differences between these animals (either their behaviour or abundance) can help determine if invasive ants are affecting them.

As with pitfall traps the idea is to compare invaded versus uninvaded sites, to get an idea of how the behaviour or abundance of other animals in the area differs. 

The species chosen will depend on the nature of the area - for example in a conservation area you might focus on a threatened species. But sometimes it is easier and simpler to compare the abundance of common species, for example sea birds, land birds, lizards and crabs.

It is easy to modify the survey methods below to look for non-target effects of a treatment programme instead of the impact of ants. Instead of comparing survey results between invaded and uninvaded areas, simply compare results from the same treatment area, before and after treatment. More information about monitoring non-target effects can be found in the monitoring section.

Visual surveys of birds

An easy method to assess the impact of ants on birds is with visual surveys. Choose at least two (more is better) observation points 25-50 m apart depending on the size of the area. Choose around four bird species to survey instead of counting every type of bird.

It is best to choose bird species that are either common, easy to identify, or endemic (of high importance because they are found only in that area) for this type of monitoring.

A person who is good at identifying birds while they are flying then stands at the observation point and counts every bird of the selected species seen over a 20 minute period. These surveys should be done at each observation point on at least 5 separate days (both before and after treatment).

Make sure to do these surveys at the same time every monitoring day. Morning is the best time because birds tend to be more active then. 

Back to top

Sea birds

Nesting seabirds such as the Akiaki (white tern) are affected by ants (© Monica Gruber)

Comparing life stages of nesting seabirds in invaded and uninvaded areas can indicate if invasive ants are having impacts on reproduction. (Please note that if this method is used to assess non-target effects of treatment on seabirds the before and after surveys should be done within one week of each other. Otherwise the effect of time on seabird development will make it difficult to tell if treatment affects development). If possible, try to assess ant abundance. For yellow crazy ants this can be done with card counts. If this isn't possible, record if the ants are present or not. Below is an example of information recorded for Akiaki (white terns) in Tokelau, comparing invaded and uninvaded areas. We recorded: ant abundance (0=none, L=low, M=medium, H=high) rat abundance (L=low, M=medium, H=high) potential nesting or display trees, and for each tree: height of nesting spot (if there was a nest) ant presence on tree presence of a single bird or pair of Akiaki only presence of egg, downy chick, pullus* or flying juvenile

The proportion of successfully hatched nesting attempts (downy, pullus*, juvenile) was compared between areas with different abundances of yellow crazy ants. Results were recorded in a table like the one below.

* A pullus is a chick at an intermediate stage when it is still not able to fly

Other factors will also affect differences in seabirds numbers, and so these should be assessed as well. These factors might include availability of suitable nesting trees (or sites for ground-nesting birds), or the presence of rats, which can also affect seabird numbers. We recorded the presence or absence of rats in the area as rats also affect birds.

We also tested yellow crazy ant preference for lures to get some idea of whether they might be attracted to sea bird eggs, or newly hatched chicks. In most cases they were more attracted to the sugar lures. But other ant species might have different preferences and preferences differ at different times of year. For example many ants require more protein when new queens and males are being produced and sometimes this is not all year round.

Various lures in plastic cup used to test yellow crazy ant potential attraction to newly hatched seabirds Left to right: honey, peanut butter, raw chicken egg (© Monica Gruber & Allan Burne)

The above approach could easily be modified for other birds also.

Back to top

Land birds Comparing life stages can work for land birds also. In addition, if the birds eat fruit, the potential impact of ants on their feeding can be assessed with model fruits! This method can also be used to assess the effect of treatment on birds' feeding. Fruiting displays can be made with wooden dowels with holes to thread wire or smaller sticks through. Model fruits are stuck on the end of them. You can make model fruits by rolling non-toxic, odourless, water-resistant modelling clay (typically red) into balls. These are pressed onto the end of the wires or sticks. The fruits should be larger than the gape (open mouth width) of native birds in the area.
The displays are attached to plants at a height of about 1.2 m. A number of displays should be placed in invaded and uninvaded areas, around 10 - 25 m apart (depending on the size of the infestations). The size of both the areas should be the same i.e. on a transect line 100 m long.		Artificial branches with artificial fruits are placed in trees (© Monica Gruber, Pacific Biosecurity)

A transect is just an imaginary line that is used to help standardise measurements of environmental data. For example, the number of species (e.g. trees) will probably be more along a line of 200 metres than 100 metres. To be able to make reliable conclusions, it's important to keep measurements as standardised as possible.

Preferably more than one transect in invaded and uninvaded areas should be chosen e.g. five transects. Make sure the transects are far enough away from each other so that you don't sample the same birds.

After a day, the number of peck marks on the fruits is counted, and compared between uninvaded and invaded sites. If ants are stopping the birds form feeding, the number of pecks will be lower in the invaded sites.

Back to top

Lizards

Lizards are a common species in the Pacific, and easy to measure (© Meghan Cooling, Pacific Biosecurity)

Lizard abundance can be assessed visually, but can also be measured using a special type of pitfall trap (not the same as the ones used for ants). Lizard pitfall traps are very different to the ant traps. They do not kill the lizard, and must be made so the lizards caught are safe from harm. The traps must be covered to protect the lizards from rain and predators, and have a raised area (stones for example) inside the trap. See the New Zealand Department of Conservation guide if you wish to try this method. Otherwise lizards can be visually counted and compared between uninvaded and invaded areas, or before and after treatment.

Back to top

Crabs

Land crabs are often harmed by invasive ants in the Pacific. The most iconic of the land crabs are the coconut crabs, a key food resource on many Pacific islands. Although food harvesting might make it difficult to assess the impact of ants on coconut crabs, their abundance can be assessed using a transect survey. The transect survey might seem complex but could be simplified to include body measurements and CPUE (catch per unit effort = total number of crabs along a transect / total number of baits used on a transect). Other, more abundant types of crabs are easier to survey than coconut crabs. For example, hermit crabs can be counted and compared between invaded and uninvaded areas, or before and after treatment.

Hermit crabs are often harmed by invasive ants in the Pacific (© Meghan Cooling, Pacific Biosecurity)

Back to top

Effects on people

It's usually obvious if people are experiencing negative effects. They'll tell you.

Often, word of mouth reports are enough to justify action, but to get funding sometimes funders can want a more objective assessment of the effects on people. For example when yellow crazy ants were at high abundance on Atafu, Pacific Biosecurity undertook a survey to see what the community thought. Their answers were very useful as a justification for action when seeking funding.

Sometimes surveys of this type need to have Human Ethics Approval - this will depend on the requirements of your employer or local regulations. But even if Human Ethics Approvals aren't required, it's important to consider what people are being asked (and the way the questions are being asked), how their privacy is maintained, and that they are fully informed about the end use of the information they are giving. Interviewees should never feel an obligation to answer any question and should be free to change their mind about participating.

Perhaps most importantly, surveys of this type help weigh up community attitudes to the invasive species relative to the possible non-target effects of the pesticides. This also helps to prioritise action.

Surveys should also be undertaken on any health effects if treatment is being undertaken in an inhabited area. Pacific Biosecurity's yellow crazy ant management work in Tokelau and Kiritimati includes health surveys, which you can use to base your own health survey on.

Back to top

Information sources

Agosti, Majer, Alonso, Schultz (eds). 2000. Ants: standard methods for measuring and monitoring biodiversity. Biological Diversity Handbook Series. Smithsonian Institution Press, Washington DC

Boland, Smith, Maple, Tiernan, Barr, Reeves, Napier. 2011. Heli-baiting using low concentration fipronil to control invasive yellow crazy ant supercolonies on Christmas Island, Indian Ocean. in Veitch, Clout, Towns (eds.) 2011. Island Invasives: Eradication and Management. Proceedings of the International Conference on Island Invasives. Gland, Switzerland: IUCN and Auckland, New Zealand: CBB. xii + 542pp

Brown, Davidson. 1977. Competition between seed-eating rodents and ants in desert ecosystems. Science 196:880–882

Davis, O'Dowd, Green, Mac Nally. 2008. Effects of an alien ant invasion on abundance, behavior, and reproductive success of endemic island birds. Conservation Biology 22: 1165-1176

Davis, O'Dowd, Mac Nally, Green. 2010. Invasive ants disrupt frugivory by endemic island birds. Biology Letters 6:85-88

Gruber, Burne, Abbott, Pierce, Lester. 2012. Population decline but increased distribution of an invasive ant genotype on a Pacific atoll. Biological Invasions: 1-14

Hare. 2012. Herpetofauna: pitfall trapping v 1.0. New Zealand Department of Conservation

Holway, Lach, Suarez, Tsutsui and Case. 2002. The causes and consequences of ant invasions. Annual Review of Ecology, Evolution, and Systematics 33: 181-233

Kaiser-Bunbury, Cuthbert, Fox, Birch, Bunbury. 2014. Invasion of yellow crazy ant Anoplolepis gracilipes in a Seychelles UNESCO palm forest. NeoBiota 22: 43-57

McNatty, Abbott, Lester. 2009. Invasive ants compete with and modify the trophic ecology of hermit crabs on tropical islands. Oecologia 160(1):187-94.

Pierce, Gruber, Atherton, Burne, Valu, Whistler. 2011. A conservation survey of Tokelau. Eco Oceania Pty Ltd Plan for Tokelau Administration and Critical Ecosystem Partnership Fund

Schiller. 1992. Assessment of the status of the coconut crab Birgus latro on Niue Island with recommendations regarding an appropriate resource management strategy. A Consultancy report prepared for The Food And Agriculture Organisation of the United Nations

Uchida, Mori, Kojima, Hayana, Sakairi, Chiba. 2016. Effects of an invasive ant on land snails in the Ogasawara Islands. Conservation Biology. DOI: 10.1111/cobi.12724

content reviewed by Phil Lester, Victoria University of Wellington, August 2016; Souad Boudjelas, Pacific Invasives Initiative, November 2016