Active surveillance on avian influenza virus in Egypt has been extensively performed on a regular basis since February 2006. For instance, our surveillance in 2009 confirmed continuous A/H5N1 infections in commercial poultry farms, backyard birds and LBMs in Egypt. Although, outbreaks in commercial poultry farms (0.1%, n = 23/22024) could be neglected as a great risk for the commercial poultry sectors in contrast to the backyards (10.5%, n = 151/1435). However, lack of compensation hampers the reporting of outbreaks and might result in a skewed/blurred picture of the actual field situation. It is worth mentioning that the current capacity of slaughterhouses in Egypt was estimated roughly to be 30 – 60% of the national meat poultry production and veterinary inspection is insufficient . Moreover, illegal trading of unexamined commercial poultry and backyard birds into LBMs is not uncommon; therefore this might explain the higher incidence of the virus in LBMs (11.4%, n = 108/944). These results are in accordance with our surveillance conducted in cooperation with the Food and Agriculture Organization of the United Nations (FAO) in 2006  and 2007 – 2008 . In contrast to our results, Kayali et al.  reported 6.8% (n = 192/2827) positivity rate in commercial farms, 3.3% (n = 34/1024) in LBMs and only 0.9% (n = 12/1381) in the backyard flocks. However, the latter group conducted a targeted surveillance from August 2009 to July 2010 in six governorates only where 53% (n = 2959/5562) of their collected samples were from poultry from two governorates.
The viral circulation in vaccinated and non vaccinated birds was previously reported; particularly during the winter seasons of 2006 – 2008 [3, 7]. However, the results obtained herein showed that the epidemiology of A/H5N1 in Egypt in 2009 has changed over time with outbreaks, especially in backyards, occurred in the warmer months of the year; spring and summer which may indicate establishment and adaptation of the virus to the environmental conditions. This observation is in accordance with findings of Kayali et al.  in Egypt in 2009–2010 and was also found in Vietnam  and in opposition to a winter-associated pattern of AIV in other countries [21–23]. In the current study, A/H5N1 was more prevalent in LBMs that had waterfowls (and/or turkeys) but not chickens alone. Unfortunately, paucity of epidemiological data is an obstacle for identification of the source of birds, particularly waterfowl, in the markets and curbs trace back of infection. Nevertheless, due to cultural factors the source of ducks in LBMs is usually the backyards while chickens are usually come from commercial farms . Also, in contrast to chickens, waterfowl can be silently infected with A/H5N1 [25–29] which may maintain the virus in the LBMs for longer periods. It has been previously described that A/H5N1 infections are high in Upper Egypt particularly in the Nile Delta which could be considered as the influenza epicentre in Egypt where major metropolitan areas with dense human populations are concentrated and a lot of poultry are likely to be traded and consumed [3, 7, 10, 13, 19, 26].
Previous comprehensive phylogenetic analyses described temporal pattern of A/H5N1 in Egypt and neither geographical nor species-linked pattern were observed [12, 13]. Interestingly, Ck18-F and Dk224-F were isolated from two different poultry farms at the same governorate (El-Qaluibia) but they belonged to two different genetic lineages (Figure 3) whereas Ck6-BY and Tk1-M isolated from LBM and commercial farms, respectively from two different provinces (about 800 km apart) clustered together within the 18.104.22.168 clade. This could be explained by the rapid and random movement of poultry nationwide and mix of different poultry and marketing sectors. The topology of the HA gene of Ck18-F isolated from commercial poultry is different from the topology of the NA gene (Figure 3) which possibly is due to reassortment. However, the full genome sequence is required to confirm this notion.
Although isolation of many AIV subtypes from wild birds in Egypt has been previously reported, we failed to identify any A/H5N1 from wild birds which may indicate, in the context of this surveillance, no role of wild birds for spread of the virus in domesticated poultry. This is also in accordance with previous negative A/H5N1 results in samples collected from wild birds in Lake El-Manzala in 2006  and 2009 . Indeed, commercial poultry-LBMs-backyards cycle in Egypt is closely integrated and any breach will eminently affect poultry and endanger public health. Therefore we suggest that enforcement of biosecurity measures should be the first line of defence while vaccination acts only as an ancillary tool for control of A/H5N1. Depopulation of infected holdings requires prompt and fair compensation. Lack thereof will severely hamper effective eradication of the disease.