Card counts

Card counts are a quick, simple and reliable way to monitoring the abundance of yellow crazy ant infestations. They give a general indication of the ants' abundance. Also, unlike other sampling methods, only invaded sites need to be checked.

The major benefit of card counts is that they provide an objective assessment of whether the ants have reached an abundance that indicates they are having effects on the environment. Thus they are an important tool for management. It would be ideal if an objective assessment like this was possible for other ant species too.

Card counts are not suitable for counting ants that regularly form trails. The method could be adapted for other ants that behave in a 'crazy' way, such as the black crazy ant (Paratrechina longicornis) and the browsing ant (Lepisiota frauenfeldi). 

Standard approach

The good thing about card counts is that they are an objective assessment of abundance that has been linked directly to the impacts of these ants. So it can give a good indication when control should be considered.

For yellow crazy ants, a threshold of 37 ants crossing a 10 x 10 cm card in thirty seconds has been identified as the point that abundance is enough to warrant management (based on the level of environmental harm by the ant). 

Choosing sites

First you need to plan where you will do your card counts. Usually this means choosing a few areas (sites).

Choose sites that are typical of the invaded area and  at least 10 m away from the sea / lagoon. On small atoll islets this should preferably be around halfway between the ocean and the lagoon.

A note on replication

Replication simply means having more than one sample of something. You need replication because there can be a lot of differences between single samples. This is variation, and can really affect your conclusions. 

You need more than one card count transect at each site. How many you choose will depend on practicalities like the total area you have to cover. 

You need to know that your results are sound. This is why you have many sites and more than one transect line within sites. This is called nested replication, and removes the effect of variation at different spatial scale (between the sites and between the lines in the transects).

At least two replicate card counts per site are recommended. The sites should be given names or numbers to identify them.


Ant activity is measured at 11 stations spaced at 5 m intervals along a 50 m transect. Three transects are assessed at a site.

The card count procedure works well if there are three people: each person can do one transect each (walking side-by-side 10 m apart) and the time taken is a lot less than one person alone.

Each person counts the number of ants crossing a white card over a period of 30 seconds.

The mean value for all counting performed on the 3 transects is the card count total. If the card count total is more than 37 then management is thought necessary.

Back to top

Three people 10 m apart place cards on the ground and count the number of ants crossing the card in a thirty second period. Each person repeats the procedure at 5 m intervals along a 50 m transect (© Allan Burne, Pacific Biosecurity)


No special equipment is needed but it is easier if closed shoes are worn (rather than open sandals or jandals).

  • one square card per person. Make a square card (laminated is preferable as it is stronger and can be re-used, but a sheet of A4 paper works OK). The card / paper should measure 10 x 10 cm
  • a watch for timing
  • pen and paper to record results

Detailed procedure

A team of four is best - three counters and one timer. But one of the counters can act as a timer if only three people (or less) are available. The timer also records the results.

One person times the three counters as they count the number of ants crossing a laminated 10 x 10 cm card (inset) over a 30 second period (© Warren Butcher, Pacific Biosecurity).
  1. Choose a starting station. This is the first point along each transect. Three people stand 10 m apart
  2. Each person clears a space on the ground large enough for the card. This is easily done by scraping with the side of a shoe
  3. Place the card on the ground
  4. The timer calls 'go' and times 30 seconds
  5. For 30 seconds the counters count the number of ants that cross the card. If possible they try not to count the same ant twice
  6. After 30 seconds the timer calls 'stop'
  7. The counters tell the timer their results and the timer records this number
  8. The counters move 5 m (about 6 paces) to the next station. Steps 3 – 7 are repeated
  9. Steps 3- 7 are repeated until 11 numbers are recorded by each counter (i.e. one for each station)
  10. The mean value for all 3 transects is the card count total for the monitoring site

One person can do this by themselves. The difference is they will have to do one transect at a time.

Back to top

Card count data should be recorded in a table like the one below. This assists in calculating the average, which will inform whether management is necessary.

Sometimes although yellow crazy ants are around, few ants choose to cross the card. This is why multiple counts are done - to even out this type of variation. It is not unusual for card counts to be 0, even if there are a lot of crazy ants around.

Also, if a card count happens to be close by a nest, the counts can be much higher than the average. In the example below station 7 had high counts for all three transects. As this covers around 20 m, this result is unlikely to be due to a single nest, especially as the counts for station 6 and 8 were also relatively high. 

counts done: 23/11/2015 4pm-6pm    
site station transect 1 transect 2 transect 3
Atafu 1 12 1 5
Atafu 2 0 1 0
Atafu 3 0 0 15
Atafu 4 0 1 0
Atafu 5 1 12 0
Atafu 6 17 42 18
Atafu 7 68 73 21
Atafu 8 45 0 25
Atafu 9 0 0 0
Atafu 10 0 1 0
Atafu 11 12 1 3
124.7 <-mean|totals-> 155 132 87

Back to top

Low yellow crazy ant abundance

When yellow crazy ants are at low abundance they can still be having negative effects. When they are in low abundance, it is not possible to use standard card counts, as the values will always be 0.

A method to monitor yellow crazy ants at low abundance was developed for the Vallée de Mai in the Seychelles. We recommend this method be used when small effects need to be detected. If you use this method, it would be good to know what the measurement is when ants are having an effect in the environment.

A grid of 10 × 10 m quadrats is surveyed.

Quadrats are spaced 75 × 75 m apart along parallel transects across the area, avoiding areas with water or on large boulders. 

Ant activity is only recorded when rain is not likely, between 8 am and 4 pm when ant activity is fairly constant.

Ant activity is measured by placing a 15 × 15 cm laminated white card on the ground or into the leaf litter with an absorbent cotton pad soaked in 15% sugar solution in the centre.

After a settling period of 3 minutes all yellow crazy ants that cross the sheet within three minutes are counted.

The number of individuals that cross each sheet per minute is referred to as ‘ant activity’.

Five ant activity counts are taken per quadrat, one at each corner and one in the centre, and the mean value of each quadrat is measured. 

Information Sources

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.). Island invasives: eradication and management. IUCN, Gland, Switzerland.

Burne, Barbieri, Gruber. 2015-2019. Management Plan Atafu, Tokelau. Pacific Biosecurity Management Plan

Green, Comport, Slip. 2004. The management and control of the invasive alien crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean: the aerial baiting campaign September 2002. Report to Environment Australia and the Crazy Ant Steering Committee

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

content reviewed by Phil Lester August 2016

Previous page: Sticky traps Next page: Assessing ant impacts