Population ecology and implications for conservation of Cleome droserifolia: a threatened xerophyte

Journal ofAridEnvironments (1990) 19, 269-282

Population ecology and implications for conservation of Cleome droserifolia: a threatened xerophyte Ahmad K. Hegazy* Accepted 28 November 1989 The population ecology of a perennialdesert plant [Cleome droserifolia (Forssk). Del.] was studied in Egypt. Seed bank, seed rain, seed dispersal, seedlingand juvenileestablishment, and adult population were examined. Complete survivorship curve, life table and fecundity schedulewereconstructed for a naturally growing population. The results were discussed in the context of species conservation. The following recommendations are essential to conserve the species and to assist in devising a management strategy: (1) at least one population should be maintained in situ for conservation, scientific interest, educationand aesthetics; (2) harvestingin the remaining populationsshould be restricted to the oldest individuals; (3) plant collectors, botanical gardens and herbaria should collectseeds; and (4) experimentalresearch is needed on in situ and ex situ propagationof the species.

Introduction Conservation and management of threatened species require quantitative data on the size, life, death, fertility and fecundity of populations. Although many data have been amassed on botanical aspects, detailed studies of the population ecology of threatened desert plants have not yet been published. Such population ecology studies may enable future predictions as well as descriptions of past events based on current status-an important asset to preserve the species. The majority of studies on plant populations have concentrated on short-lived herbaceous species. The published accounts are too fragmentary and do not cover the different facets of plant population ecology (cf. Harper, 1977; Solbrig et al., 1977). This generallack of information on the population ecology of threatened or rare species is a major problem in devising conservation measures. The available data include simple studies in the form of annual records of the numbers of individuals and numerical data on the dynamics ofa few species (Bradshaw & Doody, 1978; Griggs & Jain, 1983; Fiedler, 1987; Hutchings, 1987). The species under investigation [Cleome droserifolia (Forssk). Del. Capparidaceae] forms a coenopopulation sensu Rabotnov (1969). Because of ecological disasters such as several successive seasons with lower than average precipitation, and over-exploitation of mature plants by desert dwellers and herbalists for use in folk medicine (Osborn, 1968), the species is threatened. The study of its seed bank, seed rain, seed dispersal, seedling establishment, survival and reproduction will aid in conservation of the species. •Department ofBotany, Faculty ofScience, University ofCairo, Giza, Egypt.

A.K.HEGAZY

270

~

.

u

30

!!~

..

{I2l

.

20

~

Q.

E 10

~

Koltomio (251 m) 26'2°C 12·5mm

..

60

.... ".

............ ..

.

E E

.'":-.: :.. :.. :..:. :.. :. :. ; . :..

0·.··:·:·: . :·:.. :.. :·: . :.:.: . :.. :.: .

: : : :~: : : : : :.: : : :}: :/: : :

....................... ,," . " .

.

.

.~

.

20

"." " "

.

~

JFM4MJJ4S0ND

Month

Figure 1. Climatediagram(average of 12 years, 1976-1987) of Kattamia stationsituated closeto the study area. [Following Walter an Lieth (1960)]. Study area

Climate The study area is located in the eastern desert of Egypt (latitude 29° IS' N, longitude 32° 26' E). The climate diagram of Kattamia station, situated about 40 miles from the study area is shown in Fig. 1.

Soil The soil is rocky and consists of a mixture of angular rock fragments and loose gravel. The rocky soil protects the fine soil material (particle size less than 2 mm) which is largely confined to pockets and cracks. The fine soil consists of 58·2% coarse sand, 8'6% fine sand, 3·5% silt and 2'2% clay. The soil is calcareous and contains about 28·5% total carbonates. It is alkaline (pH 7'8), with total soluble salts as low as 0'61 %. It is very poor in organic matter (0'13%).

Vegetation About 30 perennial species constitute the plant community in the study area. The importance value (IV) for each of the species was estimated following Ayyad (1970) by calculating the density, dominance and frequency (IV = relative density + relative dominance + relative frequency). The community is dominated by Cleome droserifolia which attained an IV of 67'8 (out of a total of 300). The three species Zygophyllum decumbens, Launaeaspinosa and Gymnocarpos decandrum each attained an IV of more than 30, whilst the six speciesZygophylium coccineum, Iphiona mucronata, Crotalaria aegyptiaca, Salvia aegyptiaca, Scrophularia deserti and Lavandula stricta attained IVs of about 10. Another species in the plant community had an IV less than 10. Materials and methods

Seed population Seed bank. Fifty randomly distributed soil samples were collected in January, May, July and October. Each sample was collected from a previously moistened site 20 x 20 x 5 em deep, scraped and placed in a bag. The soil was then air dried, broken up and sieved

POPULATION ECOLOGY AND CONSERVATION OF A XEROPHYTE

271

through a 2-mm sieve to remove coarse stones and plant fragments. The seeds were then separated from the soil by flotation in a concentrated aqueous solution of potassium carbonate and left to dry in air. The settled soils were air dried and re-examined manually for any further seeds. The total seed bank was then estimated as number per m 2 • The seed bank was categorized into four components following Schafer & Chilcote (1969). 1. Pex are exogenously dormant seeds; the germinable seed bank under optimum conditions. 2. Pend are endogenously dormant seeds; these are viable seeds which failed to germinate under optimum conditions. 3. D, are the non-viable seeds among the ungerminated seeds in the Pex test. They included all seeds which showed a negative response to triphenyl tetrazolium chloride. 4. D g represents the seeds that germinate in the soil. About 1000 seeds were tested for germinability in sterile sandy soil (oven dried at 100°C for 24 h) supporting the natural population.

Seed rain. The population size and different plant size classes were estimated. Plant size was measured by the crown diameter method (Mueller-Dombois & Ellenberg, 1974). The population was then categorized into different size classes (cohorts). The number of seeds produced per plant, and the percentage contribution of each cohort to the total seed output were estimated. Seed dispersal. Ten individual plants were selected. The criteria for their selection included: (1) isolation from conspecific adults, and (2) location at different sites within the study area. The dispersal of seeds in the neighbourhood of every individual was screened using estimation of total seed bank. Squares (10 x 10em) were assigned to concentric rings (all compass directions taken together) drawn from the source plant at the centre. Seed dispersal was estimated for increments of 10 em to a distance of 100 em, and then in increments of 25 em to 300 em from the source plant. The average seed densities were estimated by dividing the sum of the counts by the sum of the area of squares included in each concentric ring. Seedling andjuvenilepopulation Within the study area, a total of 50 quadrats (1 x 1 m) were located at random. For the purpose of this study, a '0'-year seedling was defined as any individual plant developing from a germinating seed until it reaches one year old. Juvenile plants (i.e. plants failing to reach reproductive maturity) have an age range from 1 to 5 years. Monthly observations of seedling emergence and mortality were made (biweekly observations during January to March). Newly-emerged seedlings were tagged by placing a coloured plastic toothpick adjacent to the seedling. Juvenile individuals were aged by counting previous years' leaves, which remain undecomposed at the base ofthe plant. To estimate age, total leaf number was divided by mean annual leaf production. The' l'-year plants were identified by the presence of dead leaves or scars corresponding to the '0' year seedlings, the '2' year plants with dead leaves or scars corresponding to the '1' year plants, and so on through 3, 4 and 5 years.

Adult population The difficulty of estimating the age of adult plants increases progressively with the age. Accordingly, besides the age estimation, adult plants were divided into 13 size classes. Phenologic and demographic observations were then carried out, and monthly observations made on phenology. Ten individuals from every size class were marked and observed. The recorded phenophases were leafing, flowering, fruiting and seeding. The percentage of individuals failing to produce flowers was recorded within every size class.

A.K.HEGAZY

272

Life table and fecundity schedule

It was assumed that the investigated population was in a stable state, i.e. the year-to-year variation in the population stability averages out by dealing with different stages from seed to senescing individuals (Begon & Mortimer, 1986). Such an assumption may be unjustifiable for many perennial populations, but for Cleome droserifolia it may be reasonable because (1) the species forms a coenopopulation including all phases of the life cycle from seed to senescing individuals. (2) The cushion habit of the species. It was possible to sort the population into different size (age) classes. (3) The oldest age class is modestly represented in the population. It forms about 0·5% of adult individuals, and many such individuals fail to produce flowers. (4) The investigated population is isolated and there is no sign of disturbance or human impact on it. Having made these assumptions, the life table and fecundity schedule were constructed, as described by Pielou (1977) and Begon & Mortimer (1986). Results Seed population Seed bank. The mean numbers and percentages of different seed bank categories in the top 5 em of soil are shown in Table 1. There was a tendency for a significantly larger seed bank to occur around October when the total buried seed density was 3898 seeds m -2. This was true for the different seed categories, except for Pend seeds, which attained the largest density of 1813seeds m- 2 in January. The high October seed bank figures follow the peak time of seed shedding and dispersal. Seed rain. The number of fruits produced per individual plant increased with size (age) until the size class 40-45 dnr' where the largest number of fruits was produced per individual [Fig. 2(a)]. The number of seeds per individual plant per size class parallels the number of fruits [Fig. 2(b)]. The greatest seed rain was produced by cohorts which ranged in size between 1 and 10 dm", i.e, about 10 to 15 years old [Fig. 2(c)]. This was the most dense age-specific seed rain in the population. The average number of seeds per capsule was 66 ± 23 seeds. The estimated average total number of seeds produced by the whole population during the year of observation wasI9,946,360 seeds or about 4987 seeds m- 2 • Table 1. Average density of seeds (number m- 2 ) in each category of the buried seed population of Cleome droserifolia. See textfor explanation of symbols. Numbers in brackets represent thepercentage out of total buried seeds. Figures arerounded to the nearest wholenumber. Note that ripe seed shedding and dispersal starts by the end of May/early June and reached itspeak during September Month Seed

January

Pex Pend Dg Dn

0249 (06'8) 1823 (49'5) 0032 (00'9) 1574 (42-8) 3678 (100)

Total

May

July

0248 (08'4)" 1383 (47'1)" 0027 (00'9) 1284 (43'6)" 2942 (100)"

0363 (10'4) 1396 (40'2) 0021 (00'6) 1694 (48'8) 3474 (100)

Significantly lower than the average at 95% level. b Significantly higher than the average at 95% level.

a

October 0394 (lO'I)b 1551 (39'8) 0039 (01'1) 1914 (49'I)b 3898 (lOO)b

Average 0314 (08'9) 1538 (44'2) 0030(00'8) 1617 (46'1) 3498 (100)

POPULATION ECOLOGY AND CONSERVATION OF A XEROPHYTE

273

1200 1000

~

:;

800

'0

600

Q;

.0

E

400

=>

z

200

80000 (b)

70000 60000

'"

50000

'0

40000

"0

'" '" '" Q;

.0

~ 30000

z

20000 10000

4,10 6

3,106

'" '"

"0

'" '"

2 d0 6

0

73

l-

1,106

0:,..

0;;. ...., ~ "0, ~

\ \ l.1' ,/ 0;s. /0 0

tb ~lJ'

/. ~ , '" 0

\

~ ~ ~

\

'0 ~\ \

~ 'b '" '0 ~

~

°

lS'0

Size cress (dm 3 )

Figure 2. (a) Number offruits per individual plant per size class. (b) Number of seeds per individual plant persize class. (c) Total number of seeds produced per size class. Vertical lines around the mean indicate the standard deviations.

Seed dispersal. The seed density decreased with increasing distance from the source plant

(Fig. 3). The maximum dispersal distance was about 2 m from the source plant. About 87% of seeds was deposited within O'5 m of the seed source.

Seedling population The pattern of seedling dynamics was associated with the soil moisture content. The greatest losses of seedlings were incurred in summer and fall when the soil moisture

A.K.HEGAZY

274 120

100 N

E

6

. -.. u>

80

-e

~

60

0 ~

~

E 40 :::J

Z

20

0·5

1·0

1'5

2·0

2'5

Distance (m)

Figure 3. Average seed density in relation to distance from the source individual plant (all compass directions taken together). Vertical lines indicate the standard deviations.

1000 500 100 N

E

0

Q

50

u>

co

.. -.

~

"0

u>

10 10

0 ~

~

E

:::J

Z

2

a

J

F M A M J

J

A 5

a

N

a

Month

Figure 4. Survivorship curve of seedling cohort (log scale) and soil moisture content (%) throughout the year. • . , Seedlings; 0- - -0, soil moisture.

reached less than 2% (Fig. 4). Seedling numbers showed a non-significant decrease during January and February, followed by a steady decline during the rest of the year.

Adult population Phenology. Vegetative growth occurred throughout the year. Leafing began in February, but most leaf production occurred during spring and early summer. The number of green leaves per main branch reached 13 during May, falling to only three in December and January (Fig. 5). Leaf area index varied greatly during the year, ranging from 0'4 in

POPULATION ECOLOGY AND CONSERVATION OF A XEROPHYTE

., "

275

3

"0

C

.,

C

-., C C

...J

15,---------------,

JFMAMJJAS

Month

~

Vegetative

IZJ Flowering/fruiting

•

Seed dispersal

~

Yellowing of leaves

Figure 5. Phenological spectrum, number of green (e e) and dying-back leaves (0- - -0) per branch, and leaf area index (square metre leaf area per square metre plant surface area) throughout the year.

January to 2·9 in May-June in parallel with leaf numbers. Flowering started in April and extended to October, with an average period of flowering per individual plants of 4 months. Fruiting pattern was parallel to the flowering activity. About 5 to 30% of young individuals and from 5 to 20% of old plants did not flower during the year of observation (Table 2). All individuals within the size classes from 10 to 40 dm" were flowering.

Size distribution. The size (age) distribution of the population is shown in Fig. 6. Up to

82·5% of total adults were between 0'1 and 10 dm", The largest size class (greater than 50 dm") represented only about 0·5% of the total adult population. The spatial distribution of Table 2. Percentage 0/individuals o/Cleome droserifolia that/ailed toflower during

theyear 0/observation

Size class (dm'')

0'1-0'5 0'5-01·0 01-05 05-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45

45-50 >50

Percentage

30 15 10 05 00 00 00

00

00 00

05 05 20

A.K.HEGAZY

276

individual plants as estimated by Hopkins' index (aggregation = 4·8 clumped pattern of the population (Pielou, 1977).

± 3'9) indicated the

Life table andfecundity schedule The life table for the standing population is given in Table 3. The semi-logarithmic plot of Fig. 7 indicates the original survivorship (Ix) as unity. The curve shows extremely high seed and seedling mortality. This was followed by linear mortality with age up to 20 years, then by a plateau of low mortality in adults between 20 and 60 years old. The curve shows a truncated decline for individuals older than 60 years. In the life table (Table 3), there were more individuals in their fourth year. The same was true for individuals in their 36th and 55th years. There were therefore negative deaths (d x values) and meaningless mortality rates. Mean expectation of life (ex) was highest in the juvenile stages and lowest in seed and seedling stages. A fecundity schedule was constructed based on seed production data (Table 4). The estimation of age-specific fecundity (d x ) shows that while the age-specific fecundity was increasing, the size of the older population cohorts (ax) was gradually declining. The estimated net reproductive rate (R a) as shown in Table 4 was 0'99996. This value suggests a nearly 'stable' population level rather than a negative population growth. The attained intrinsic rate of increase per capita per year (-1.33 x 10- 6 ) seems to be within the acceptable limits of the experimental error. The distribution of the contributions to R o during age intervals Clxbx values) for different cohorts of the adult population are given in Fig. 8(a). The reproductive value (Vx) of the different cohorts are plotted in Fig. 8(b). The largest values of Vx are distributed over the intermediate cohorts of the population. Discussion

Seedpopulation From the seasonal sampling, it was apparent that a persistent seed bank was maintained of which about 8'9% was exogenously dormant, 44'2% endogenously dormant and 46'1% non-viable. Exogenous dormancy mechanisms delay seed germination to periods when environmental conditions are suitable for successful seedling establishment. Endogenous 30

V>

"3

20

f--

,.-

"0

-

o

"6

~ '0 <>!!

I-10

l1-n~ o~q,\

'o~U'

...,./

!S'\-o'6'tb~U'..;~~~"k

-o~~\

is'

~o\\

o"'",o':i-'O~o

\~o

Size class (9m3)

Figure 6. Age distributionofthe adult population. The individuals are groupedinto 13sizeclasses. The heightof a columnshows the percentage of all plants that make up the givenclass.

Adult plants

Seed rain Seed bank Germinable seed in soil Seedlings Juvenile plants

Stage

00'1--00'5 00·5--01-0 01'0--05'0 05'0-10'0 10·0-15·0 15'0-20·0 20'0-25'0 25'0-30·0 30·0-35'0 35'0-40'0 40'0-45·0 45·0-50·0 >50

-

0 01 ± 0'3* 02 ± 0·5 03 ± 0·8 04 ± 1 05 ± 1 06±2 08 ± 3 12 ± 3 14±4 21 ± 5 25 ± 4 30 ± 6 36 ± 5 43± 7 55 ± 7 61 ± 8 64± 8 78 ± 11

x

Cohort tdm")

410 260 86 73 46 58 25 28 19 17 11

446

19946360 13992000 119000 12562 787 766 610 755 642 594

ax 1·000 0'702 5·97 x 10- 3 6'30 x 10- 4 3·95 x 10- 5 3·84 x 10- 5 3·06 x 10- 5 3·79 x 10- 5 3·22 x 10- 5 2'98 x 10- 5 2·24 X 10- 5 2'06 x 10- 5 1'30 x 10- 5 4·31 x 10- 6 3'66 x 10- 6 2'31 X 10- 6 2·91 x 10- 6 1·25 x 10- 6 1'40 x 10- 6 9'53 x 10- 7 8'52 x 10- 7 5·52 X 10- 7

I"

-

0'298 0'696 5'34 X 10- 3 5·91 X 10- 4 1'10 X 10- 6 7·80 X 10- 6 -7'30 X 10- 6 5'70 X 10- 6 2'40 X 10- 6 6'50 X 10- 6 1·80 X 10- 6 7'60 X 10- 6 8·69 X 10- 6 6·50 X 10- 7 1-35 X 10- 7 -6,00 X 10- 7 1'66 X 10- 6 -1,50 X 10- 7 4'47 X 10- 7 1'01 X 10- 7 3'00 X 10- 7

dx

-

0'319 0'106 0'352

-

0·571

-

0·150 0'075 0'218 0·080 0'369 0'669 0·151 0'037

-

0'298 0'992 0'895 0'938 0·028 0·227

q"

1'21 0'51 0·65 0·96 6·76 5'94 6·31 3·22 3'70 2'95 2·77 1'99 1-85 3·56 3·30 4·00 2'28 3'46 2'21 2'11 1'11 0·50

ex

Table 3. Life table calculated for Cleome droserifolia at the sample sitefrom data collected on the standing population, x = estimated age(years), ax = numbers per ageclass, lx= proportion oforiginal cohort surviving to thestartof eachage, dx = proportion oforiginal cohort dyingduring each age, qx = mortality rate, and e; = expectation offuture life. * = standard deviation.

N

..... .....

tTl

>-l

-<

:r:

'"d

0

::d

>< tTl

>

't1

0

Z

>-l 0

~ >

-

tTl

r:J)

Z

0

o

0

Z

-< >

0

0

t'""

0

tTl

o

Z

(5

> >-l

c: r-

'"d

'"d

0

A.K.HEGAZY

278

Seeds

10- 3

Seedlings

a.

:.c

'" o

.:': >

::>

Ul

10

20

30

40

50

60

70

80

Age (years)

Figure 7. Survivorship curve (logscale) of Cleome droserifolia.

dormancy, by producing a diverse seed population, prevents synchronous germination which could make the population vulnerable to extinction. This suggests that increased environmental uncertainty in deserts due to drought and high temperature stresses has resulted in increased seed dormancy (Harper, 1977). This was also predicted by Cohen (1966) and Levins (1969). Generally, the dormant seeds rely on environmental cues which indicate the probability in the future of conditions favourable for growth (Amen, 1968; Karssen, 1980). The persistent seed bank reflects the history ofthe population and may enable species to survive environmental changes (stresses) which may kill seedling, juvenile and/or adult populations (Thompson, 1976; Baskin & Baskin, 1978). Although the number of seeds per individual plant increases with plant size [Fig. 2(b)], the total number of seeds per size class was significantly higher in the two size classes 1-5 and 5-10 dm" [Fig. 2(c)], due to the larger number ofindividuals in these two size classes. The larger size classes tended to show a significant pattern of alternate low and high number of seeds every year per size class, indicating irregular fecundity as age increases. Most of the seeds remained within 1 m of the source plant. This short-distance dispersal is due to non-violent capsule dehiscence so that the seeds fall close to the source plant. The combination of high seed production with short distance dispersal means that most seeds will land in a spot likely to be favourable for germination and seedling establishment, because the parent plant was able to establish there (Harper, 1977). This dispersal pattern may explain the clumped dispersion of the adult plants.

POPULATION ECOLOGY AND CONSERVATION OF A XEROPHYTE

279

Seedling, juvenile and adultpopulations The seedling and juvenile stages suffered high mortality relative to the adult plants. Only about O'35% of the total seed population had the potential to germinate in soil. Out of these germinable seeds, about 10% may give seedlings. About 28% of the seedlings survived to produce juvenile plants, and 58% of these reached reproductive maturity. This increased survivorship with age is easily explained by the development of deeper roots, with less dependence upon surface moisture (Naylor, 1985). The critical role of seedling and juvenile establishment in determining the subsequent dynamics of the adult population lies in its importance as a stage of natural selection, as the high mortality is likely to eliminate low fitness genotypes (Cook, 1979). This is supported by other results (e.g. Stebbins, 1974; Symonides, 1977; Meagher & Thompson, 1987) where survival of seedlings and juveniles varies from year to year. Despite declining adult populations with age, the population protected from human impact and other disturbances may persist. This is because there are sufficient individuals in seedling, juvenile and young adult size classes to replace the old individuals, as well as the uninterrupted regeneration and phenological cycle.

Life table andfecundity schedule The results of the life table (Table 3) are summarised as a survival curve (Fig. 7). The curve exhibits a combination oftype I, II and III curves (Deevey, 1947). The high mortality of seed and seedlings coincides with the type III curve. The linear mortality with age in juvenile and adults up to 20 years old indicates a fairly constant mortality rate (type II curve). The low mortality of adults above 20 years old, which was followed by truncated decline for individuals older than 60 years can be described as a type I curve. A similar survival curve was exhibited by the palm Euterpe globosa from tropical forests (Van Valen, 1975). Meanwhile, different curves were exhibited for other species (Sarukhan & Harper, Table 4. Fecundity schedule calculated for Cleome droserifolia based on seed production. x = estimated age(Years), ar = numbers perage-class, l ; = proportion of original cohort surviving to the startof each age, br = average number of seeds per individual, Vr = reproductive value,R o = netreproductive rate, T= generation time, and r = intrinsic rate of increase

x 06 08 12 14 21 25 30 36 43 55 61 64 78

ar 594 446 410 260 86 73 46 58 25 28 19 17 11

Ix 2·98 x 2'24 x 2'06 x 1·30x 4'31 x 3'66 x 2'31 x 2'91 x 1'25 x 1'40 X 9·53 x 8'52 x 5'52 x

10- 5 10- 5 10- 5 10- 5 10- 6 10- 6 10- 6 10- 6 10- 6

10- 6 10- 7 10- 7 10- 7

bx

i»,

xlrbr

Vr

00730 02191 08765 11886 13612 21314 30079 39840 43691 51194 66201 63279 51925

0'02175 0·04908 0'18056 0'15452 0'05867 0'07801 0·06948 0'11593 0'05461 0'07167 0·06309 0'05393 0'02866 IO'99996

0'13050 0'39264 2'16672 2·16328 1·23207 1'95025 2·08440 4'17348 2-34823 3-94185 3-84849 3'45152 2'23548 I30.11891

033557 044425 004413 058415 140672 147664 203259 137349 217521 161413 154307 105465 051925

R o = 0'99996; r = -1'33 x 10- 6 ; T = 30'12.

A.K.HEGAZY

280 0·20

(a)

0·15

..,,' ~,

0·10 0'05 0·01 25xl0 4

(b)

20, 104

15, 10 4 ."..

IOxl0 4 5' 10 4

Figure 8. (a) Distribution of the contributions of net reproductive rate based on seed production during age intervals(lxbx values). (b) Reproductiverate (Vx) for the different sizeclasses. 1973; Leverich & Levin 1979; Hutchings, 1987). To sum up, Cleome droserifolia displayed a combination of Deevey's three curves as stage follows stage in its life cycle. The estimation of mean expectation offuture life (ex) for the various cohorts surviving to successive years (Table 3) showed that it was highest among juveniles, while declining in the adult population. This may explain the relatively long period of juvenality (about 5 years) before establishment of adult individuals. The value of R, was 0'99996, suggesting a nearly 'stable' population level for the time being, rather than 'negative' population growth. This indicates that, in spite of the relatively high reproductive values (Vx), the great fecundity of individuals at various size classes does not lead to a 'positive' population growth. This is due to the high losses during seed, seedling and juvenile stages. Only one in 57,135 seeds of Cleome droserifolia yields an adult individual of average 6 years old, which is the earliest reproductive age. In comparison, Van Valen (1975) found that one in 432,000 seeds of the palm Euterpe globosa yields an adult that attains an age of 51 years which is the first reproductive age.

Implications for conservation What are the implications of the population ecology of Cleome droserifolia for its conservation? What is the future of this species? The persistent seed population provides a natural way in which stock can be maintained. It would ensure re-establishment of the population even if no reproductive individuals produced a seed crop in several consecutive years (Baskin & Baskin, 1978). Secondly, it increases the genetic variability and stability of the population, because seeds of different ages germinate every year (Epling et al., 1960; Gottlieb, 1974). Thirdly, an understanding of the population dynamics of buried viable seeds provides a basis for regeneration techniques of the species. Fourthly, the knowledge of the range of environmental requirements for seed germination is important for

POPULATION ECOLOGY AND CONSERVATION OF A XEROPHYTE

281

evaluation of artificial seeding for species propagation, a fact not investigated in this study. Fifthly, the presence of high levels of seed dormancy may prevent all seeds from responding to occasional unpredictable showers which may stimulate germination but do not supply enough moisture for seedling establishment and growth. How can such a threatened species be utilized efficiently without causing its decline or even extinction? The challenge is not to prevent people from using it, but to rationalize the harvest while maintaining the population. Since the reproductive value increases to a peak and then drops with age, the collection of plants should be restricted to the oldest individuals of the population. This solution may be applicable for the cropping of many wild plants, where the concept of preventing people from using any natural resource is not easily acceptable, particularly in Third World countries. Under the current conditions of uncontrolled exploitation of many populations and a low rate of recruitment to the adult populations, the future of Cleome droserifolia in the wild is jeopardized. The following recommendations towards a conservation strategy are put forward: (1) some populations ofthe species should be conserved in situ and protected from human activities; (2) harvesting should be restricted to the oldest individuals in the population; (3) since vegetative propagation seems negligible, the route to population increase lies in creating conditions which are conducive to flowering, seed setting and seedling establishment; (4) experimental research is needed on in situ and ex situ propagation as well as seed storage. The advice and suggestions of Drs Michael Hutchings (University of Sussex), Mohamed Kassas (University of Cairo) and Walter Moser (University of Alberta) on the first draft ofthe manuscript are highly appreciated. The comments of an anonymous reviewer have cleared many ideas. I thank Drs Gamal Fahmy and Salama Ouf for kind help in the field work, and Mr Mahmoud Sayed who accompanied me in some field trips. References Amen, R. D. (1968). A model of seed dormancy. Botanical Review, 34: 1-31. Ayyad, M. A. (1970). Applications of the point-centred quarter method to the vegetation of two types of desert habitat at Mareotis. United ArabRepublic Journal of Botany, 13: 225-234. Baskin, J. M. & Baskin, C. C. (1978).The seed bank in a population of an endemic plant speciesand its ecological significance. Biological Conservation, 14: 125-130. Begon, M. & Mortimer, M. (1986). Population Ecology: A Unified StudyofAnimals andPlants. (2nd Edn) New York: Sinauer Associates. 220 pp. Bradshaw, M. E. & Doody, J. P. (1978). Plant population studies and their relevance to nature conservation. Biological Conservation, 14: 223-242. Cohen, D. (1966). Optimizing reproduction in a randomly varying environment. Journal of Theoretical Biology, 12: 119-129. Cook, R. E. (1979).Patterns of juvenilemortality and recruitment in plants. In: Solbrig, O. T., Jain, S., Johnson, G. B. & Raven, P. H. (Eds), Topics in PlantPopulation Biology. pp. 207-231. New York: Columbia University Press. 589 pp. Deevey, E. S., Jr. (1947). Life tables for natural populations of animals. Quarterly ReviewofBiology, 22: 283-314. Epling, C., Lewis, H. & Ball, F. M. (1960). The breeding group and seed storage: A study in population dynamics. Evolution, 14: 238-255. Fiedler, P. L. (1987). Life history and population dynamics of rare and common Mariposa lilies (Calochortus pursh. Liliaceae). Journal of Ecology, 75: 977-995. Gottlieb, L. D. (1974).Genetic stability in a peripheral isolate of Stephanomeria exigua spp. coronaria that fluctuates in population size. Genetics, 76: 551-556. Griggs, F. T. & Jain, S. K. (1983). Conservation of vernal pool plants in California. 11- Population biologyof a rare and unique grass genus Orcuttia. Biological Conservation, 27: 171-193. Harper, J. L. (1977). Population Biology of Plants. London: Academic Press. 892 pp, Hutchings, M. J. (1987).The population biologyof the early spider orchid, Ophrys sphegodes Mill. 1A demographic study from 1975to 1984. Journal of Ecology, 75: 711-727.

282

A.K.HEGAZY

Karssen, C. M. (1980/81). Environmental conditions and endogenous mechanisms involved in secondary dormancy of seeds. IsraelJoumal of Botany, 29: 45-64. Leverich, W. J. & Levin, D. A. (1979). Age-specific survivorship and reproduction in Phlox drummondii. American Naturalist, 113: 881-903. Levins, R. (1969). Dormancy as an adaptive strategy. In: Dormancy and Survival. Symposium of the Society for Experimental Biology, 23: 1-10. Meagher, T. R. & Thompson, E. (1987). Analysis of parentage for naturally established seedlings of Chamaelirium luteum (Liliaceae). Ecology, 68: 803-812. Mueller-Dombois, D. & Ellenberg, H. (1974). Aims andMethods of Vegetation Ecology. New York: John Wiley. 547 pp. Naylor, R. E. L. (1985). Establishment and peri-establishment mortality. In: White, J. (Ed.), Studies on Plant Demography. pp. 95-109. London: Academic Press. 393 pp. Osborn, D. J. (1968). Notes on medicinal and other uses of plants in Egypt. Economic Botany, 22: 165-177. Pielou, E. C. (1977). Mathematical Ecology. New York: John Wiley. 385 pp. Rabotnov, T. A. (1969). On coenopopulations of perennial herbaceous plants in natural coenoses. Vegetatio, 19: 87-95. Sarukhan, J. & Harper, J. L. (1973). Studies on plant demography: Ranunculus repens L., R. bulbosus L. and R. acris L. 1- Population flux and survivorship. JournalofEcology, 61: 675-716. Schafer, D. E. & Chilcote, D. O. (1969). Factors influencing persistence and depletion in buried seed populations. 1- A model for analysis of parameters of buried seed persistence and depletion. Crop Science, 9: 417-419. Solbrig, O. T., Jain, S., Johnson, G. B. & Raven, P. H. (Eds) (1979). Topics in Plant Population Biology. New York: Columbia University Press. 589 pp. Stebbins, G. L. (1974). Flowering Plants Evolution Above the Species Level. Cambridge: Belknap Press. 397 pp. Symonides, E. (1977). Mortality of seedlings in natural psammophyte populations. E kologia Polska, 25: 635-651. Thompson, P. A. (1976). Factors involved in the selection of plant resources for conservation as seed in gene banks. Biological Conservation, 10: 159-167. Van Valen, L. (1975). Life, death and energy of a tree. Biotropica, 7: 260-269. Walter, H. & Lie,t,h, H. (1960). Klimadiagrammweltatlas. Iena: WEB Gustav Fischer Verlag.