Abstract
This study aimed to: (1) Establish if rodent populations in sewers are separate from surface ones (Sewer Interceptor study) (2) Investigate the numbers and distribution of rats in sewers (Preferred Location study / Long term study/ Hot spot study) (3) Identify Rat behaviours that might affect numbers and control measures (Bait trial / Preferred Location study). To achieve these aims sewer rat populations were monitored in the London sewer systems over a 20-year period. More detailed studies on rat behaviour were carried out in sequestered sewer systems. The Interceptor study showed that rats are not confined in the system by water traps and cannot be viewed as a separate population from surface rats. They can, especially during periods of low flow, swim underwater, upstream, in the dark, to access or exit the system. The Preferred Location study showed that rats like to find and use dry locations within the sewer network. There are distinct sets of behaviours, which occur in dry locations and not in wet ones and vice versa. The Bait trial confirmed other studies showing that sewer rats are neophobic and this factor needs to be taken into account when dealing with them. The more rats there were, the shorter the avoidance period. In the experiment here, the mean time until consumption decreased from 8.5 for one rat to 5.5 days for three rats. The Long Term study found that the trend in rat numbers has been in decline over the whole twenty-year period of the study in the location studied. In the first thirteen years of the study, the proportion of bait takes decreased significantly from approximately 0.17 in 1986/87 to 0.03 in 1998/99 (F=17.75, P<0.001: df 1,11). Since then, the decline has continued with an exponential curve best fitting the data. The Hot Spot study showed that Rats are not evenly distributed throughout the sewer network but occur, year on year in "Hotspots". These were locations, which showed up to 17 times more than the statistically significant level of rodent activity. This activity occurred in the same location year after year although not throughout the entire study period. These locations were statistically significant (p<0.0001 after using the Bonferroni correction).