PDF《Review article》

Benton, 2006;

Goulson Article published by EDP Sciences

368 P.H. Williams, J.L. Osborne et al., 2008). This review di?ers by consider- ing evidence world-wide for extinction and de- cline and discusses the factors that threaten the persistence of bumblebee species and the fac- tors that in?uence which species are most sus- ceptible. We note which species world-wide have become most threatened and which ar- eas of the world support the largest numbers of threatened species (we do not consider the par- asitic species of the subgenus Psithyrus, be- cause information is too sparse). We then dis- cuss conservation-management strategies and ?nish with suggestions for future work. 2. EVIDENCE FOR DECLINES Most concern about bumblebee conserva- tion has come from widespread perceptions that many bumblebee species have declined in local abundance and in their range extent (i.e. through local extirpations), without go- ing globally extinct. The dates and rates of these changes are often very uncertain, be- cause baseline data for comparisons are rarely available. Data collected 50C100 years ago using standardised and repeatable sampling methods for measuring abundance are almost completely lacking, but fortunately there are data on range extent that can provide informa- tion. The earliest worries about changes in the status of bumblebee species were expressed in Britain by Free and Butler (1959), who wrote that It is commonly supposed that the bumble- bee population has declined in recent years . Some species had been completely lost from Britain. For B. pomorum (Panzer), there had only ever been a few records from the south- east coast of Britain before the last record in 1864, so that it may never have been a long- term part of the fauna (Alford, 1975). For B. cullumanus (Kirby), females were always rare and especially di?cult to identify, making it unlikely to be recorded, although the more easily recognized males were recorded regu- larly from the southern chalk grasslands until '

c. 1941'

(Williams P.H., 1982). Now it ap- pears that B. subterraneus may have joined this list of losses (Edwards and Williams P.H., 2004). Although once described as common and even locally abundant (Sladen, 1912), it subsequently became less common (Yarrow in Free and Butler, 1959) or scarce (Alford, 1975), but was then seen regularly in only low numbers at Dungeness on the south-east coast (Williams P.H., 1989b), until the last individ- ual was found in a pitfall trap in 1988. Since then, repeated visits have found no more indi- viduals. Among all British species, strong regional range reductions seemed to be apparent in the maps for the

1960 onwards data collected by the Bumblebee Distribution Maps Scheme (BDMS: Alford, 1980). This pattern becomes less sensitive to heterogeneity in local sam- pling e?ort (Williams P.H., 2000) when the data are plotted as maps of regional range using large area units (Williams P.H., 1982, 2005). For example, Figure

1 compares the de- clines in regional species richness (although it cannot show any e?ects of range expan- sions), showing the strongest declines in cen- tral England. Of the

16 currently persisting non-parasitic bumblebee species in Britain (excluding B. magnus Vogt and B. cryptarum (Fabricius), which cannot yet be recognised reliably in both sexes and castes without re- course to molecular methods, Murray et al., 2008), more than half have been suggested to be rare and in decline (Williams P.H., 1986;

Goulson, 2003;

Edwards and Williams P.H., 2004;

Goulson et al., 2005;

Benton, 2006). These authors agree on declines for B. distinguendus Morawitz, B. humilis Il- liger, B. muscorum (Linnaeus), B. ruderatus (Fabricius), B. subterraneus, and B. sylvarum (Linnaeus). Other species that have been sug- gested to be rare or declining are B. jonel- lus (Kirby), B. monticola Smith, B. ruderar- ius (Müller), and B. soroeensis (Fabricius). Although there is evidence that these latter species remain widespread in at least some re- gions (Williams P.H., 1985;