The sample area was in Khyber Pakhtunkhwa, Pakistan, with latitude of 34.9526 °N and a longitude of 72.3311 °E ( Figure 1 ), with a total geographical area of approximately 101,741 km².For the phytosociological data collection, thirty-nine (39) sites were selected across three elevational gradients (high elevation = (AL 3) > 1420 m a.s.l; medium elevation = (AL 2) > 1256 m a.s.l; and low elevation = (AL 1) > 331 m a.s.l). The field survey was conducted from February 2018 to May 2020. The research locations’ latitude, longitude, and aspect degree were recorded using a GPS, and the slope was assessed using a clinometer. The average annual precipitation was 400 mm, with the lowest and highest temperatures of 2°C and 40°C, respectively. Within the thirty-nine permanent quadrats (5×5 = 25 m²) used to measure biomass, they found that members of the species were healthy, disease-free, and devoid of physical injury. A composite soil sample was collected from each site from the plant’s root zone, i.e., 50-80cm depth. The samples were sorted into three categories for floristic data based on altitudinal zonation and delivered to the Swat Agriculture Research Institute (ARI). Soil extract was used to determine Organic matter, Potassium (K), Nitrogen (N), and Phosphorus (P) (Rhoades, 1993; Nelson and Sommers, 1996). Available water was calculated by using an online calculator developed by Saxton and Rawls (2006).

Each quadrat was tagged with different colours to calculate the density of C. procera (i.e., the number of individuals per unit area). The leaf area was measured using the trace paper method, and the leaves/branches were randomly counted by picking five separate branches. Similarly, the standard procedure was used to compute plant diameter, height (m), and volume (Pandey and Singh, 2011). Plant height and diameter ratio fluctuate in different microhabitats. Hence these characteristics were considered when estimating shrub size (Shaltut and Ayyad, 1990). To calculate dry biomass, 39 complete plants were dug up, and each portion (stem, leaf, and root) was preserved individually in polythene bags. The destructive approach was used to calculate the plant biomass (Wang et al., 2008). All of the samples were oven-dried for 48 hours at 65 degrees Celsius. The dry weight was obtained by weighing the samples after being returned to the Department of Botany at the University of Malakand in KP, Pakistan. Based on the data of the plant crown diameter (C), we calculated the canopy area (CA m²) and volume (V m³) using the techniques given by Zeng et al. (2010).

Where CA stand for canopy area, V for volume and H for height of studied plant.

A regression equation was developed in order to determine the association between shrub total biomass, used as the dependent variable, and other growth characteristics such as diameter, height, height-to-diameter ratio, canopy area, and volume (D², H, D²H, CA, and CV), used as the explanatory factors.

To determine whether the regression equations were accurate representations of the data, the coefficient of determination (R²), standard error estimate (SEE), and F-values between the pairs of x and y variables were calculated (Zeng et al., 2010).

Analysis of variance (ANOVA) and Tukey HSD post hoc tests was employed to identify significant differences in the biomass ratio and the other variables evaluated. To estimate growth parameters and biomass, an allometric regression equation was applied. In order to determine if the biomass and other growth metrics exhibited significant responses, a permutational multivariate analysis of variance (PERMANOVA) was performed, and ordinations were formulated. This approach is thought to be particularly well adapted to ecological data, especially when working with a small sample size (Andringa et al., 2019). The “Adonis” technique in the vegan package 1.13-2 was utilized in the PERMANOVA with the Euclidean distance approach (Sabo et al., 2019; Oksanen et al., 2020). Non-metric multidimensional scaling (NMDS) was used to depict the ordinals, utilizing the default parameters of vegan’s “meta MDS” method (Kruakal, 1964). A total of 999 permutations were employed in this investigation. The statistical analysis was carried out using R software using the dplyr, tidyverse, ggplot2, and vegan packages.

Table 1 summarises the spectrum of characteristics and biomass changes of Calotropis procera throughout the three altitudinal groups. There were no significant differences in total biomass between various slope locations and slope angles. Plant populations may be distinguished by physical traits, such as stem height (2.63 ± 0.21 m) and crown diameter (2.18 ± 0.27 m) for those found at high elevations (AL 3). C. procera grows well in disturbed and open environments. Hence individuals reported from rural regions with low altitude (AL 1) were shorter (H=1.66 ± 0.13 m, D=1.99 ± 0.13 m) ( Table 1 ). The H/D ratio was larger than one in all habitat categories, with urban sites with high altitude (AL 3) have the greatest ratio (1.21 ± 1.95 m) and rural sites having the lowest (0.83 ± 0.73 m). There was a considerable fluctuation in the height of the stem, diameter, size index, and volume of C. procera in the current study, indicating the species’ growth adaptability in disturbed habitats. The biomass of the investigated plant was high at a high altitude (79.85 ± 4.90) in the sample locations but exhibited heterogeneity in all groups.

The standing effect of ground factors (slope, elevation, slope direction) on the biomass of C. procera shrubs and soil variables (N %, P mg/Kg, K mg/Kg, OM %, and AW %) using boxplots ( Figure 2 ). The result declared that the soil variables (OM %, N %, P mg/Kg, K mg/Kg) and biomass of plant show a decreasing trend from low to high altitude except for available water contents (AW %) and Potassium (K mg/Kg) contents. Figure 2 demonstrates that different stand factors significantly affect the distribution pattern, whereas when slope direction changes from north-facing slope (AS 1) to south-facing slope (AS 2), no other significant differences are noticed except for available water (AW %) and nitrogen content (N%). Furthermore, it had a significant increasing trend of slop from 0-5° to 15°-20° for all soil variables ( Figure 2 ).

The influence of stand variables on various biomass and morphological features varied greatly, as shown in Table 2 . The Euclidean distance for the distribution pattern of C. procera biomass and morphological traits was used to determine the PERMANOVA findings in this research. There were significant differences in all site parameters on biomass (p< 0.005). The biomass by site parameters had a considerable degree of explanation (R²) (p< 0.05). A greater impact of slope direction and elevation was observed on C. procera shrub biomass, such as for biomass, altitude (0.45)< aspect-degree (0.62)< slope-aspect (0.76); for height, altitude (0.73) > aspect-degree (0.39)< slope-aspect (0.73); for canopy area, altitude (0.63) > aspect-degree (0.25)< slope-aspect (0.82) and canopy volume, altitude (0.33)< slope-aspect (0.50)< aspect-degree (0.87). The volume was significantly affected by the interplay of height and aspect degree. The biomass distribution patterns of C. procera were significantly affected by the interaction of the three site variables. From Table 2 , it can be seen that Euclidean distance can be used to calculate the PERMANOVA results for individual stand factors.

The biomass partitioning pattern of C. procera shrubs was mapped with environmental factors using the non-metric multidimensional scaling (NMDS) approach based on Euclidean distance. As illustrated in Figure 3 , the distribution of loci within the two-dimensional NMDS classification scheme of C. procera shrub biomass can be seen. The NMDS sorting process may provide evidence of the link between shrub biomass and stand variables through limiting the number of axes and allowing the sorting axes to reflect particular environmental gradients. The findings of the PERMANOVA were confirmed by the NMDS ordination (stress = 0.04%), which demonstrated clustering of topographic effects on biomass allocation patterns, with the highest divergence between the variables ( Figure 3 ). There were some changes in the distribution pattern of C. procera shrub biomass on various slope orientations and altitudes. The slope direction had the most significant changes.

Table 3 shows the equation parameters, accuracy, and goodness equations established for the C. procera biomass estimate. C. procera was estimated using allometric equations across the three elevation groups. H, D, D²H, CA, and CV were used to see which growth characteristics predicted biomass most accurately. Based on linear modeling, the best fit equations for biomass estimate were derived using H, D, and CV as best predicted. These equations were statistically significant (p< 0.01). The linear equations showed the highest R²-value for flower biomass and total biomass except for the foliar biomass. Linear models with V, CA and H as independent variables showed the best fit for biomass of stem, leaves, and flower, respectively at rural areas ( Table 3 ). However, D² as independent variable had lowest R² for total biomass. On the other side, at urban area stem biomass have high R² value with respect to other organ biomass while at roadside flower biomass have high R² value with V as independent variable.

In contrast, stem biomass is linked to diameter and height at low elevations in urban and roadside environments ( Table 3 ). The accuracy measurements suggest that one or two factors may adequately describe C. procera biomass ( Table 3 ). There is, nevertheless, room to improve the model’s explanatory power by including more independent variables. D², D²H, CA, and H are employed to analyse shrub biomass regressions as independent variables. The linear model was the best fit for providing a common statistical foundation for characterizing size–biomass connections in this investigation. H, D²H, and CV were the most relevant independent variables for this C. procera investigation. The linear regression model was used to show a significant relationship between the aboveground biomass in response to different environmental variables such as altitude, aspect degree, slope aspect, N %, P mg/Kg, K mg/Kg, OM %, and AW % ( Figure 4 ).