0198566123.pdf

The data quality problems seen for the plants
undoubtedly afflict even the best of the animal data
sets for Krakatau, but are less well understood.
Here we take them essentially at face value. The
butterfly data (Fig. 5.10) and land-bird data
(Fig. 5.11) show similar trends coincident with for-

est closure, but contrasting trends towards the end

of the data series, with butterflies seemingly climb-

ing more steeply and birds approaching an asymp-

tote. In the first two decades, the poverty of the

vegetation arguably presented a limited array of

opportunities for butterfly species, many of which

KRAKATAU—SUCCESSION, DISPERSAL STRUCTURE, AND HIERARCHIES 139

60

(a) (b)

40

20

1883 1903 1923

Year

Cumulative

Immigration

Extinction

Number of species

Nrec

1943 1963 1983 1883 1903 1923

Year

1943 1963 1983

0

Species per annum

2.0

1.0

0

Figure 5.10Butterfly recolonization data for Rakata Island (Krakatau) 1883–1989 (redrawn from Bush and Whittaker 1991). The peaks in
immigration and extinction appear to correlate with habitat succession as grassland gave way to forest habitat on the islands. (a) Cumulative
total and number of species recorded at each survey (Nrec), as Fig. 5.10. (b) Immigration and extinction curves, as Fig. 5.8.

60

50

40

30

20

10

0

80

(a) (b)

70

60

50

40

30

20

10

0

SLRLAWRWA Ferns Wind

Primary dispersal mode

Percentage of species

absent in 1979–94

Number of species

Sea Animal

1 island

Present

Absent

2 island

3 island

Abundance category

Figure 5.9Persisting and failing species from the 1883–1934 flora of Krakatau (Rakata, Sertung, and Panjang islands), as determined by
survey data for presence on the islands between 1979 and 1994. (a) Presence/absence in relation to abundance categories derived from
descriptive accounts from earlier botanical work, where Ssingleton records; LRlocalized and rare; LAlocalized but abundant;
WRwidespread but rare (low density); WAwidespread and abundant. (b) Losses from the flora in relation to primary dispersal mode and
the number of islands on which each species was recorded in the 1883–1934 period. Wind, sea, and animal refer to spermatophytes only; ferns
are all wind-dispersed, in the form of spores. (Source: Whittaker et al. 2000, Figures 2 and 3). Unsurprisingly, initially rare species have a lower
rate of persistence than initially widespread species. Further scrutiny of the raw data allows exclusion of a number of the extinctions documented
in this figure and a lower overall estimate of extinction rate in the flora. Such an adjustment, if made to Fig. 5.8b, would induce a steeper
negative trend to the extinction curve than shown there.