**3.1.4 Species accumulation curve**

Species accumulation curve for the random plots showed different flattening levels in vegetation types of both the zones. The species accumulation as a function of number of individuals is given in Fig. 3a. Deciduous forest in both the zones showed a clear tendency towards flattening of the curve; however, zone-2 shows further increment due to its high species richness. Contrastingly, moist deciduous in both the zones doesn't show a typical flattening compared to the deciduous system. It seems that moist deciduous forest need to

Fig. 3a. Species accumulation vs. number of individuals in Southern and Northern zones of Eastern Ghats, Andhra Pradesh

Fig. 3b. Species accumulation curve for the six transects in Northern and Southern Eastern Ghats of Andhra Pradesh

be characterized and analysed based further on its altitudinal and climatic conditions to derive a species-area relation. The species-accumulation curves for the continuous plots of the six sites are given in Fig 3b. Site 1 and 4 were initially steep, and later we observed a tendency towards flattening and similar such pattern was observed for the Site 5 & 6. Site 2 & 3 didn't reach an asymptote due to high species richness and as well landscape heterogeneity.

### **3.1.5 Stand density and basal area**

160 The Dynamical Processes of Biodiversity – Case Studies of Evolution and Spatial Distribution

represented families include Fabaceae (101), Euphorbiaceae (58), Rubiaceae (51) and Acanthaceae (48). Dominant families with respect to number of individuals are Euphorbiaceae (1786), Mimosaceae (1669), Combretaceae (1606), Fabaceae (1577), Rubiaceae (1368) and Caesalpinaceae (1282). Families with rare occurrences represented by single and

For the continuous plot of the six sites, a total of 57 families belonging to 2,457 individuals were recorded from the present study. Taxonomically well-represented families include Rubiaceae (18), Euphorbiaceae (16), Fabaceae (11) and Caesalpiniaceae (9). Dominant families with respect to number of individuals are Mimosaceae (337), Combretaceae (274), Fabaceae (191), Rubiaceae (154) and Caesalpinaceae (138). Families with rare occurrences represented by single and double species were 36 for both the study sites. Top ten families explain the species characteristics and found to be 66% (1620 individuals out of 2,457

Species accumulation curve for the random plots showed different flattening levels in vegetation types of both the zones. The species accumulation as a function of number of individuals is given in Fig. 3a. Deciduous forest in both the zones showed a clear tendency towards flattening of the curve; however, zone-2 shows further increment due to its high species richness. Contrastingly, moist deciduous in both the zones doesn't show a typical flattening compared to the deciduous system. It seems that moist deciduous forest need to Northern Eastern Ghats (Zone-1) Southern Eastern Ghats (Zone-2)

Fig. 3a. Species accumulation vs. number of individuals in Southern and Northern zones of

20 410 800 1190 1581 1971 <sup>0</sup>

16 666 1315 1965 2614 3264 3913 No. of Individuals

double species were 66 for the whole study area.

individuals) dominant for the study site.

**3.1.4 Species accumulation curve** 

Dry Deciduous Forest

Estimate no. of species

Eastern Ghats, Andhra Pradesh

> 16 173 331 488 645 803 960 No. of Individuals

15 166 317 467 618 769 920 1071

Moist Deciduous Forest

Estimate no. of species

Stand density and basal area for the random plots in both the zones of deciduous forest is higher compared to other forest classes (Table 2). In zone-1 of deciduous forest stand density was much higher (486 stems ha-1 and 19.04 m2ha-1 basal area), when compared to zone-2 having 386 stems ha-1 and 14.67 m2ha-1 basal area. Moist deciduous forest of the zone-1 & 2 are in the range of 257-370 stems ha-1. Similarly such patterns were also observed in dry evergreen forests (266-334 stems ha-1). Degraded forest in zone-1 was relatively higher (449 stems ha-1) due to the selective felling for its timber extraction when compared to zone-2. Thorn forest in both the zones is having least stand density class compared to other types.

Stand density and basal area (BA) for the continuous plots in *site* 1, 2 and 3 is higher compared to *site* 4, 5 and 6 (Table 2). *Site* 2 has high stand density of 1050 stems ha-1 (BA 17.61 m2ha-1) as compared to *site* 3 having high basal area of 24.42 m2ha-1 (990 stems ha-1). Least basal area is seen in *site* 4 (4.35 m2ha-1) and least stand density in *site* 5 (614 stems ha-1). This is due to the dry deciduous forest in zone-2 having less density and less biomass. The girth class distribution pattern for all the sites except *site* 3 and 5 is basically "L" shaped for

Spatial Patterns of Phytodiversity - Assessing Vegetation Using (Dis) Similarity Measures 163

explain most of the explainable variation in species richness in multiple regressions with

The mrpp() function of package vegan (Oksanen *et al.*, 2007) for the R Statistics System has been implemented to evaluate whether the plots of the two zones can be attributed to different vegetation types. The two zones are clearly distinct in their vegetation composition: The observed delta is significantly lower (0.89) than the expected delta (0.93, as determined by permutation) although the difference is not very large. Furthermore A = 0.036. A is a chance-corrected estimate of the distances explained by group identity. It can be compared to a coefficient of determination of a linear model (R2). Thus, it shows that the grouping into the two zones based on species composition is not very clear. With other words they are less distinct in species composition than the NMDS suggests (Fig. 8, upper left panel). Therefore another, more robust test has been employed as well. The function anosim [vegan] provides basically the same test but acts on ranks instead of the original data. It reports an R of -0.28\*\*\*: The similarity among plots of one zone is significantly

Because the zones showed considerable grouping in the NMDS, further NMDS plots were drawn for each zone separately, to evaluate whether the species composition could than be more clearly attributed to the membership to categories of fragmentation, disturbance and richness. Fig. 10 shows that this is not the case. When the zones are considered separately no

Fig. 5. Richness (species number inside plots) is significantly lower in highly fragmented plots. With disturbance there is no influence. The categorization is very coarse. Note, that

significance was tested with simple anova (for fragmentation: F = 0.13\*\*, for disturbance: F = 0.83ns). Inference regarding the difference between the classes was obtained with

there the numbers of plots involved in the categories differ considerably. Overall

pairwise t-Tests (α = 0.05). Bonferroni correction was applied.

clear groupings according to the categories of the mentioned variables occur.

backward selection) do not show a clear pattern in the NMDS plots (Fig. 8).

higher than the similarity between plots of different zones.

the total number of individuals (i.e. G ≥10cm) (Fig. 4). We observed girth class of 10-30cm was having high number of individuals (39%). These species itself doesn't support larger biomass and density as a result there is very less representation in girth class above 200cm.

of Andhra Pradesh
